Jessica J. Moon, MD and Cynthia Mattox, MD

Anti-Vascular Endothelial Growth Factor in Filtering Surgery

Glaucoma is one of the leading causes of irreversible blindness worldwide.¹ The term describes a group of diseases that are characterized by a specific optic neuropathy,² with the strongest risk factor for development being elevated intraocular pressure (IOP). IOP is maintained by the production of aqueous humor with drainage through the trabecular meshwork in the anterior chamber angle or the uveoscleral outflow pathway. Glaucoma therapy is targeted at reducing IOP and slowing the progression of glaucoma. Conventional first-line therapy in glaucoma involves topical medications or laser trabeculoplasty. When this conservative therapy fails, glaucoma surgery can be performed to control the IOP.³

Trabeculectomy

Trabeculectomy was first described in 1968 by Cairns and has remained the gold standard of glaucoma surgery.⁴ It is a surgical procedure in which a guarded fistula is created to allow aqueous humor to drain from the anterior chamber into the subconjunctival space. The goal is to bypass the trabecular meshwork, allowing the aqueous humor to exit through the formed bleb to provide a controlled way of lowering IOP. It is the most effective procedure for reducing the IOP in glaucoma.^5,6 However, failure of trabeculectomy due to excessive postoperative scarring remains a significant challenge. Unlike in other surgeries, a completely healed wound is not desirable in a trabeculectomy. Subconjunctival scarring at the site of the bleb can cause adhesions to the episcleral tissue and lead to resealing of the bleb, ultimately preventing filtration and leading to an elevated IOP.7

Wound healing in trabeculectomy is driven by 2 key pathways, fibroblasts and angiogenesis, and both are modulated by vascular endothelial growth factor (VEGF). Tenon fibroblasts are the main effector cells in scar formation. After surgery, Tenon fibroblasts proliferate, causing excess collagen and elastin deposition.⁷ Histologic studies have shown that maximum proliferation of subconjunctival fibroblasts occurs on the third to fifth postoperative day.⁸ High VEGF levels are linked to increased fibroblast proliferation, leading to increased collagen deposition and subsequent encapsulation of blebs.⁹ Angiogenesis is another important aspect of wound healing in trabeculectomy. New vessels need to grow into the center of the healing site to bring in nutrients and supply oxygen. These vessels bring in mediators that support the rapid growth of new cells to replace the injured cells and thus promote scar deposition. It is known that increased bleb vascularity is associated with a poor prognosis in trabeculectomy.¹⁰

Current Therapy in Trabeculectomy

The current standard of modulating scar formation in trabeculectomy targets the fibroblast pathway and not the angiogenesis pathway. This therapy involves using antifibrotics to inhibit fibroblasts in wound healing. The 2 main agents used are 5-fluorouracil (5-FU) and mitomycin C (MMC). The agent 5-FU was first introduced in the early 1980s as an antimetabolite that interfered with pyrimidine metabolism. It blocks a crucial step in thymidine nucleotide synthesis and inhibits DNA replication, leading to cell death.¹¹ MMC was first introduced in 1986 as an antibiotic agent derived from Streptomyces caespitosus.¹² Unlike 5-FU, MMC is technically not an antimetabolite but an antifibroblast. It is capable of inhibiting DNA replication, mitosis, and protein synthesis therefore inhibiting fibroblasts.¹¹ Clinicians tend to favor MMC because of its prolonged effect when compared with 5-FU.

The use of these antifibrotics has improved the success of trabeculectomies.⁷ Unfortunately, even with these treatments, the “Tube Versus Trabeculectomy” study found that the probability of failure in patients with prior surgery treated with trabeculectomy with MMC was 46.9% after 5 years of follow-up.¹³ Furthermore, a major drawback of antifibrotics is their nonspecific mechanism of action. 5-FU and MMC work well because they inhibit fibroblasts, but they also have the unintended consequence of causing widespread death of other cells. This leads to a high complication rate involving cystic avascular blebs, infection, leakage, and hypotony with maculopathy. Infections range from blebitis to endophthalmitis.^14–16

These side effects have led to a search for other options with an increasing focus on the role of VEGF in wound modulation. In 1994, before commercial VEGF antibodies were available, angiogenesis inhibitors were found to have inhibitory effects on fibroblast proliferation and migration in vitro.¹⁷ Since then, more evidence has appeared showing that anti-VEGF agents inhibit wound healing.^18,19 In fact, delayed wound healing is listed as a warning on the manufacturer’s prescribing information for bevacizumab (Avastin).

Multiple studies have shown that aqueous VEGF levels are consistently elevated in eyes with glaucoma. This was first shown in 1998; the aqueous mean concentration of VEGF was higher in eyes with primary open-angle glaucoma (POAG) and higher yet in those with neovascular glaucoma (NVG).20 The source of VEGF production in glaucoma is not clear. VEGF is expressed and produced by the corneal endothelium, iris pigment epithelium, retinal pigment epithelium, ganglion cells, astrocytes, Müller cells, and choroidal fibroblasts.²¹ Not much is known about the production of VEGF in the conjunctiva or Tenon’s, but VEGF is elevated in the Tenon’s as well as in the aqueous. The size of VEGF (7.56 nm) is smaller than the pores of the trabecular meshwork and the openings of Schlemm’s canal (0.5 μm to 0.76 μm), so VEGF may be diffusing from the aqueous cavity into the subconjunctival space even in the absence of a filtration bleb. VEGF is not elevated in plasma so it is postulated that VEGF is produced by local ocular tissue.²²

It is also unknown how glaucoma upregulates VEGF but elevated IOP may have a role in increasing VEGF production. One study used rat models with induced ocular hypertension to demonstrate elevated VEGF levels in the retinal ganglion cells.²³ Another group demonstrated that aqueous VEGF levels were increased after trabeculectomy in rabbit models and remained elevated until day 30.⁹ The baseline elevation of VEGF found in the aqueous and Tenon’s of patients with glaucoma may predispose them to an even further degree of scarring after trabeculectomy.

Animal models have shown that anti-VEGF agents suppress the upregulation of VEGF after trabeculectomy.⁹ A single intraoperative injection of bevacizumab into the subconjunctival space and anterior chamber lowered VEGF levels compared to control eyes on postoperative day 4. There was a reduced density of blood vessels at the filtration site, decreased total area of collagen with trichrome stain, and a larger bleb area in the treated eye. There were no notable differences in IOP, however.⁹ Another group examined 42 trabeculectomy rabbits randomized to receive 7 subconjunctival injections of either 1.25 mg bevacizumab, 5 mg 5-FU, or balanced salt solution. The bevacizumab group had an improvement in bleb survival as well as a larger bleb height and area with less scarring on histological examination. However, they were also unable to show any changes in the IOP outcome, which is a major limitation of these animal models.²⁴

Which Anti-Vascular Endothelial Growth Factor Agent to Use

There are various isoforms of VEGF that can be targeted. In vitro analysis after trabeculectomy showed that VEGF₁₂₁ and VEGF₁₆₅ increased vascular endothelial cells whereas VEGF₁₈₉ increased fibroblast growth.²⁵ This indicates that VEGF₁₂₁ and VEGF₁₆₅ may have a larger role in blood vessel growth while VEGF₁₈₉ may have a larger role in scar formation. Presumably, an anti-VEGF that targets all these isoforms may be more effective in wound modulation than an agent that targets just one isoform.

This theory was tested by comparing pegaptanib, an anti-VEGF selective for VEGF₁₆₅, and bevacizumab, a nonselective anti-VEGF agent. Working in vivo with human and rabbit Tenon fibroblasts and in vitro with rabbit models, they found bevacizumab to be more effective, suggesting the importance of inhibiting both neovascularization and fibrosis, rather than neovascularization alone. Further studies have moved on to focus either on bevacizumab or ranibizumab, both nonselective VEGF inhibitors. At this point, it is unclear which anti-VEGF agent is better. Ranibizumab is a mature antibody that was designed to have a significantly stronger binding affinity than bevacizumab.26 However, the half-life of bevacizumab is longer than that of ranibizumab. The half-life of intravitreal bevacizumab in the vitreous is 4.32 days while the half-life of ranibizumab is 2.88 days.^27,28 In comparison, the half-life of pegatanib is 3.91 days.²⁹

Bevacizumab

There have been several randomized, controlled trials looking at the use of bevacizumab as an adjunctive treatment in trabeculectomies (Table 16-1). The results have been variable and inconclusive. The studies also varied in the route of administration and timing of the administration with no clear consensus on the better choice.

Route of Administration

The best route of administration is unknown. Topical, subconjunctival, and intravitreal administration of bevacizumab has been studied in rabbits.³⁰ The highest concentration of bevacizumab in the ocular tissues was detected with the intravitreal route and the lowest concentration with topical administration. However, subconjunctival injection was found to have a longer half-life than intravitreal injection, thought to be due to the high scleral permeability of bevacizumab causing a longer sustained-release mechanism. The only studies thus far looking at intravitreal bevacizumab with trabeculectomies have been in patients with NVG.

Subconjunctival Bevacizumab

The majority of studies examining bevacizumab have used subconjunctival bevacizumab intraoperatively during trabeculectomy.^8,31–35 A randomized, double-blind, controlled clinical trial of 38 patients compared intraoperative MMC alone, 3 subconjunctival injections of bevacizumab without MMC, and 1 application of a bevacizumab-soaked sponge intraoperatively. The bevacizumab-soaked sponge had no advantage over MMC; however, the subconjunctival injections of bevacizumab were equally effective as MMC in reducing the IOP but with fewer side effects; the MMC group had one case of bleb leak and one case of hypotony maculopathy while the bevacizumab group did not have any.³¹

Another prospective trial compared 34 eyes randomized to receive either 1 dose of 2.5 mg subconjunctival bevacizumab intraoperatively or 1 dose of MMC intraoperatively. Bevacizumab was effective but not as effective as MMC in controlling IOP (34% vs 56% reduction of IOP, respectively). Additionally, the bevacizumab group required more antiglaucoma medications after surgery for IOP control.³²

In a more recent study, MMC in conjunction with subconjunctival bevacizumab was found to be superior to bevacizumab alone. Forty-two patients with POAG were randomized to receive 2.5 mg subconjunctival bevacizumab intraoperatively with or without MMC application. Seventy-one percent of eyes in the MMC group met a target IOP of 12 mmHg without antiglaucoma medication while only 33% of eyes in the bevacizumab-alone group met the target at the 1-year follow-up. Interestingly, this study reported success rates at a much lower IOP than set by other studies that report outcomes at IOP less than 18 mmHg or 21 mmHg, which may not be adequate for advanced glaucoma.36

There have been 2 studies looking at intracameral bevacizumab administered intraoperatively during trabeculectomy surgery. A prospective clinical trial of 71 patients with either POAG or pseudoexfoliation glaucoma were randomized to receive 1.25 mg intracameral bevacizumab or intracameral balanced salt solution (BSS) at the end of the surgery. No antifibrotics were used in either group. In this study, the bevacizumab group had a higher success rate of 83% compared to the placebo group, which had a success rate of 48.5%. Unfortunately, the bevacizumab group also had a higher rate of bleb leak of 34% while the placebo group had only a 9% bleb leak rate. The study was limited by short-term follow-up of only 10 months.37

Another prospective, placebo-controlled trial of 138 patients also randomized patients to receive 1.25 mg intracameral bevacizumab or intracameral BSS at the end of surgery. In contrast, both groups also had MMC applied intraoperatively. At 1 year follow-up, the postoperative IOP was lower than baseline in both groups, but there was no significant difference between the 2 groups. The bevacizumab group, however, did have a higher success rate of 71% vs 51% in the placebo group; success was defined as at least 30% reduction in IOP from baseline, with an IOP ≤18 mmHg and >5 mmHg and no loss of light perception vision. Patients who received bevacizumab also needed fewer IOP-lowering interventions such as needling revision postoperatively.³⁸

Topical Bevacizumab

An observational case series in Poland examined 21 eyes with increased vascularity after undergoing MMC augmented trabeculectomy 9.8 ± 4.7 days previously. All patients were given standard steroid therapy and topical 5 mg/5 mL bevacizumab drops administered 5 times daily for 20.9 ± 9.8 days. Around 75% of eyes recovered with decreased injection of the bleb. However, 14% of patients also developed an early bleb leak.³⁹ For comparison, bleb leaks after trabeculectomy with MMC are reported to occur in 5% to 14.5% of cases;^40,41 bleb leaks after trabeculectomy without the use antifibrotics are relatively uncommon, ranging from 0% to 3%.^41,42

Timing of Administration

Just as the best route of administration has not been established yet, the best time to administer anti-VEGF agents is also unclear. Many of the previously mentioned studies looked at intraoperative application of bevacizumab with variable benefits. However, a few studies have looked at the postoperative use of bevacizumab in needle bleb revisions.

One study of 58 patients with failed trabeculectomy and ExPRESS shunt blebs requiring needling revisions received subconjunctival MMC in addition to either 1 mg subconjunctival bevacizumab or BSS injected posterior to the bleb after needling. There was no difference in success rates and IOP between the control and treatment groups 6 months post-procedure. However, the authors reported that the groups receiving bevacizumab had significantly less vascularity as well as broader, more diffuse blebs.⁴³

A different study comparing needle bleb revision with 5-FU to bevacizumab found 5-FU to be more effective, resulting in a statistically significant higher success rate compared with bevacizumab.33

Ranibizumab

There have been only 2 studies using ranibizumab for trabeculectomies (Table 16-1). One examined 10 patients in a prospective, open-label study that compared 0.5 mg intravitreal ranibizumab in addition to MMC use during trabeculectomy to MMC alone. There was no significant difference in IOPs between the 2 groups; however, the group receiving ranibizumab had a broader bleb area and lower vascularity at 6 months. This study was limited by a small sample size and short follow-up period.⁴⁴

The other study was a prospective, open-label trial of 24 patients undergoing trabeculectomy. One group received 0.5 mg subconjunctival ranibizumab and the other group received MMC application. The group treated with ranibizumab had a higher mean IOP at one month but there was no difference in IOP at 1-year follow-up. More patients in the ranibizumab group required additional glaucoma surgery.⁴⁵

These studies have not shown superiority of anti-VEGF agents over the current standard of antifibrotic agents. There are still questions about optimal dose, which anti-VEGF agent to use, best route of administration, as well as timing of application. Further research will need to be conducted in order to elucidate the benefit of anti-VEGF therapy in trabeculectomies.

Anti-Vascular Endothelial Growth Factor in Neovascular Glaucoma

Neovascular Glaucoma

Neovascular glaucoma (NVG) is a notoriously aggressive glaucoma that is challenging to treat and has a poor visual prognosis.⁴⁶ NVG was first described by Weiss et al⁴⁷ in 1963 as a type of glaucoma associated with new iris and angle vessels causing an increase in IOP.⁴⁷ It is a secondary glaucoma that is almost always associated with ischemic ocular conditions; 36% of all cases arise from central retinal vein occlusion, 32% from proliferative diabetic retinopathy, and 13% from ocular ischemic syndrome.⁴⁸ Retinal ischemia stimulates angiogenic factors like VEGF that cause new vessels and fibrous tissue to grow over the iris and eventually into the angle.⁴⁶ Neovascularization characteristically progresses from the pupil margin to the angle.⁴⁹

Three stages of NVG have been described. The first stage is rubeosis iridis in which there is neovascularization of the iris (NVI) with or without neovascularization of the angle (NVA) in the setting of normal IOP. In stage 2, the angle appears open yet fibrovascular tissue blocks the aqueous drainage system. Finally, in stage 3, contraction of the fibrovascular membrane causes the iris to pull over the trabecular meshwork with varying degrees of peripheral anterior synechiae. This leads to secondary angle closure and further IOP elevation.⁴⁹

It is known that hypoxia can trigger VEGF production leading to angiogenesis. VEGF concentration in the aqueous is 40 times higher in patients with NVG than in POAG and 113 times higher than in patients with cataracts.20 These elevated VEGF levels in NVG have interesting and very important implications for the use of anti-VEGF therapy in NVG.

Current Standard of Therapy

NVG is extremely difficult to manage. Medical control of IOP with topical aqueous suppressants is challenging and the effects are only temporizing. Ideally, patients with ischemic conditions who are at high risk for developing NVG should be followed closely and examined carefully at each visit looking for signs of neovascularization of the iris and angle. If any NVI or NVA is seen, anti-VEGF intravitreal injections can be performed followed by panretinal photocoagulation (PRP). PRP is the current gold standard of therapy for NVG for long-term control after anti-VEGF treatment.²⁰ It destroys ischemic retinal tissue and thus reduces the oxygen demand in the retina and subsequently the stimulus for angiogenesis.

If performed early enough during the neovascular process, even in the absence of anti-VEGF agents, PRP can cause visible regression of the vessels both posteriorly and anteriorly. Studies have shown reduced levels of VEGF in patients after PRP.⁵⁰ While PRP can be effective for early-recognized NVG, there are multiple drawbacks to relying on PRP alone. Specifically, the effects are delayed, so patients may continue to have pain and elevated IOP. PRP also causes death to healthy retinal cells and subsequent constriction of the visual field. It may also be difficult to perform because of a diminished view in patients with media opacities like vitreous hemorrhage, cataracts, or corneal edema from increased IOP. While PRP may be effective in NVG when the angle is open, its benefit in interrupting the process is less so after there is angle closure.⁵¹ Once the fibrovascular membrane extends across the angle and creates peripheral anterior synechiae, NVG will develop with permanent synechial angle closure causing severe IOP elevation in all but the most ischemic eyes, which paradoxically may have reduced aqueous suppression. At that point, anti-VEGF agents and PRP will not reverse the process and further surgical intervention with glaucoma surgery is needed.⁵¹

Once medications have failed to control IOP and synechial angle closure is present, glaucoma-filtration surgery is the usual surgical option for eyes with salvageable vision. Glaucoma-drainage implants like the Baerveldt, Molteno, or Ahmed valve aqueous drainage shunts have traditionally been preferred because their success is not as affected by the inflammation present in NVG.⁵² Trabeculectomy can be difficult to manage because the inflammation increases the risk of bleb failure. Antifibrotic agents have increased the success rate of trabeculectomy in NVG but there are still challenges.⁵³ Filtration surgery with either method can be difficult because of the extremely friable neovascular vessels that can bleed even with minor manipulation, causing hyphema and clot especially if no anti-VEGF treatment has been given preoperatively. This not only creates difficulty intraoperatively because of an impaired view but can also lead to postoperative complications with the formation of blood clots surrounding the glaucoma drainage implant tube or sclerostomy.⁵⁴

Injection of an anti-VEGF agent causes regression of the newly formed vessels in NVI and NVA. This regression of anterior segment neovascularization is grossly visible on slit lamp exam and gonioscopy and has been demonstrated in multiple case series.55–59 Neovascularization can be significantly diminished within the first 48 hours^58,59 and improvement seems to be maintained anywhere from 4 to 10 weeks.^60,61 Regression of visible signs of neovascularization ranges from 70% to 100% depending on the study.^59,62–64

There are several studies evaluating the effect of intravitreal bevacizumab in NVG on IOP. In a prospective, randomized clinical trial of 26 eyes, NVG eyes treated with 3 monthly intravitreal injections of bevacizumab had a significant reduction of IOP at 1, 3, and 9 months.⁶⁵ Most of the studies were unable to truly examine the isolated effect of intravitreal bevacizumab as many of the eyes also received concurrent PRP.^61,62,65,66

The mechanism of IOP lowering by anti-VEGF therapy in NVG is not completely understood. The most likely explanation is that regression of vessels and decreased permeability of the vessels reduces the early fibrovascular membrane covering the angle and improves drainage through the trabecular meshwork. Once the angle is closed with anterior synechiae, the only IOP effect may be a result of some remaining functional trabecular tissue that may improve with regression of newly formed vessels.⁶⁷

A study of intravitreal bevacizumab in open- vs closed-angled NVG found that efficacy may depend on the extent of synechiae formation and fibrovascular membrane contraction. While intravitreal bevacizumab benefited those with NVI and an open angle, eyes with closed-angle NVG had regression of the neovascularization without a change in IOP and 93% of the patients needed surgery within 2 months following injection.⁶²

Combining anti-VEGF treatment with PRP was studied in a retrospective comparison of 23 eyes treated with either same-day intravitreal bevacizumab with PRP to PRP alone. The benefits of combination therapy included a higher frequency of neovascular regression (100% of eyes in the combination group vs 17% of eyes in the PRP-only group) and a more rapid rate of regression (complete regression seen by 12 days in the combination group vs 127 days in the PRP-only group). The IOP was also more significantly reduced in the combination group compared to the PRP-only group. Furthermore, fewer eyes in the combination group needed further surgical intervention compared with the PRP group, although they were followed for only 143 days and 118 days, respectively.⁶⁸ Another similar study of 14 eyes confirmed the findings that neovascularization regressed more rapidly in the combination group; however, they did not find a significant difference in mean IOP between the groups. Additionally, it was discovered that despite similar angle anatomy at baseline between the 2 groups, the group that received combination therapy with bevacizumab had significantly more open-angle structures at the 6-month mark and this result persisted at 1 year.⁶⁹

Shunting devices are extremely useful for refractory cases of NVG because of their efficiency and lower rates of postoperative complications when compared to trabeculectomy in a severely inflamed eye. The choice between the various aqueous drainage devices such as Baerveldt, Ahmed, or Molteno shunts depends mostly on the surgeon’s preference. The treatment of NVG continues to be a challenge even with these devices. In a study evaluating 16 eyes with NVG treated with an Ahmed valve, the concentration of VEGF in failed cases (n = 8) showed significantly elevated levels of VEGF when compared with successful cases.70 This outlines the important role of VEGF in the successful outcome of surgery and has implications for anti-VEGF therapy with valve placement.

Multiple studies have evaluated the use of intravitreal bevacizumab in Ahmed valve implantation^71–75 (Table 16-2[A]). To date, there are no studies looking at the role of anti-VEGF therapy in other shunting devices such as the Baerveldt or Molteno for NVG. All reported studies used 1.25 mg intravitreal bevacizumab differing only in the administration either preoperatively or intraoperatively.

Preoperative Bevacizumab

The results of these studies have been very inconsistent. Two groups showed some benefit of preoperative intravitreal bevacizumab over cases that underwent implantation of an Ahmed valve alone.^71,72 However, other studies demonstrated no difference in success rate or IOP after Ahmed valve placement,^73,74 although there were significantly reduced rates of hyphema.

Another group combined preoperative intravitreal bevacizumab before Ahmed valve placement with concurrent PRP at the time of injection. This triple therapy was compared with patients receiving PRP and Ahmed alone. The results showed that patients treated with intravitreal bevacizumab had a significantly reduced IOP with a 95% success rate compared to a 50% success rate when bevacizumab was not used.⁷⁵ Other reports state PRP is the most important intervention in decreasing the need for subsequent surgery, as it significantly reduces the need for further surgery whether or not intravitreal bevacizumab is used.⁷⁶

Intraoperative Bevacizumab

Most studies have not been able to demonstrate a significant improvement in success rate of Ahmed valve surgery with the addition of intraoperative intravitreal bevacizumab. One group found a statistically significant lower IOP at 12 and 15 months, but there was no improvement in surgical failure rates between the 2 groups; failure included the need for additional surgery, light perception vision, and IOP greater than 21 despite additional medication.⁷⁷ Another study showed no difference in success rate, IOP, and visual acuity. Of note, the group that received intraoperative intravitreal bevacizumab was using fewer antiglaucoma medications than the control group and had more frequent complete regression of rubeosis iridis (80%) compared to the control group (25%).⁶³

Stay updated, free articles. Join our Telegram channel

Join