Skip Navigation LinksHome > October/November 2009 - Volume 18 - Issue 8 > In Vivo Corneal Endothelial Safety of Intracameral Bevacizum...
Journal of Glaucoma:
doi: 10.1097/IJG.0b013e3181996ed2
Original Studies

In Vivo Corneal Endothelial Safety of Intracameral Bevacizumab and Effect in Neovascular Glaucoma Combined With Ahmed Valve Implantation

Shin, Jae Pil MD; Lee, Ji Woong MD; Sohn, Byung Jae MD; Kim, Hong Kyun MD; Kim, Si Yeol MD

Free Access
Article Outline
Collapse Box

Author Information

Department of Ophthalmology, Kyungpook National University College of Medicine, Daegu, Korea

Reprints: Jae Pil Shin, MD, Department of Ophthalmology, Kyungpook National University College of Medicine, ♯50 Samduck-2 Ga, Jung-gu, Daegu 700-721, Korea (e-mail:

Received for publication January 30, 2008; accepted December 8, 2008

Collapse Box


Purpose: To evaluate the corneal toxicity of intracameral bevacizumab in rabbit eyes and also the effects of intracameral bevacizumab in neovascular glaucoma patients combined with Ahmed glaucoma valve implantation.

Materials and Methods: Intracameral bevacizumab (1.25 mg/0.05 mL) was injected into 5 eyes of 5 New Zealand white rabbits. A balanced salt solution (0.05 mL) was injected intracamerally in another 5 rabbits as a control. The corneal thickness, endothelial cell counts, and intraocular pressure were measured before and at 1, 2, and 4 weeks after the intracameral injection of bevacizumab and a balanced salt solution. Scanning electron microscopic examinations were performed at 1 and 4 weeks after injection. In the clinical cases, 6 patients who received an intracameral injection of bevacizumab (1.25 mg) and Ahmed glaucoma valve implantation were enrolled. The visual acuity, intraocular pressure, and regression of iris rubeosis by iris fluorescein angiography were recorded before and at 1 week, 1 month, and 2 months after the intracameral bevacizumab injection and Ahmed glaucoma valve implantation.

Results: No morphologic and functional changes in the corneal endothelial cells were observed in the rabbit eyes after the intracameral bevacizumab and balanced salt solution injections. Iris rubeosis regression was observed in all eyes within 1 week after the injection. The visual acuity remained stable or improved, and the intraocular pressure was controlled in all eyes throughout the follow-up period. There were no ocular and systemic complications associated with the combined procedure.

Conclusions: An intracameral injection of bevacizumab may be a safe alternative treatment for iris rubeosis in neovascular glaucoma when combined with Ahmed glaucoma valve implantation.

Neovascular glaucoma (NVG) is a devastating and serious consequence of many ischemic retinal diseases causing severe loss of vision and ocular pain. Retinal ischemia stimulates the release of various angiogenic factors, of which vascular endothelial growth factor (VEGF) is a well-known key molecule in the pathologic ocular neovascularization that induces neovascularization and vascular leakage.1 Elevated levels of VEGF in the ocular fluid had been reported in patients with NVG and other ischemic retinal diseases.2–4

The treatment of NVG is directed in 2 ways.5 One is to treat the underlying disease process, such as retinal photocoagulation, to reduce retina ischemia and inhibit the release of angiogenic factors. The other is to control the intraocular pressure (IOP) and inflammation by a medical treatment, cyclodestructive procedure or drainage device surgery. The implantation of an Ahmed glaucoma valve (AGV) for the treatment of refractory glaucoma, such as NVG, has shown promising results in controlling the IOP6–8 but not with respect to the regression of neovascularization. Therefore, it is reasonable to use pharmacologic treatments directed against the angiogenic factors in conjugation with drainage device surgery to induce the regression of neovascularization and control the IOP.

Bevacizumab (Avastin, Genentech) is a recombinant, full length, anti-VEGF monoclonal antibody that binds to all forms of VEGF-A. It was originally approved for intravenous use to treat metastatic colon cancer. The off-label use of bevacizumab in ischemic retinal disease and ocular neovascularization has shown promising results, and the regression of iris neovascularization (INV) has been reported after an intravitreal injection of bevacizumab.9–17 The pharmacokinetics and safety profiles of intravitreal bevacizumab in the retina have been reported recently.18–20 However, little is known about the corneal toxicity and effectiveness after an intracameral injection of bevacizumab. Therefore, this study evaluated the corneal toxicity of bevacizumab in rabbit eyes after an intracameral injection, and also its effectiveness in regressing new vessels when injected intracamerally combined with AGV implantation in NVG patients.

Back to Top | Article Outline


Corneal Toxicity in Rabbit Eyes

Ten New Zealand white rabbits were divided equally into 2 groups. All animal experiments were carried out in accordance with the Association for Research in Vision and Ophthalmology statement for the Use of Animals in Ophthalmic and Vision Research. This study was approved by the Laboratory Animal Care and Use Committee of Kyungpook National University College of Medicine.

The rabbits were anesthetized with intramuscular injections of ketamine hydrochloride (50 mg/kg) and xylazine hydrochloride (20 mg/kg) before each examination. The intracameral injection of bevacizumab (1.25 mg/0.05 mL) was performed in the right eye of 5 rabbits. The same dose of balanced salt solution (0.05 mL, BSS, Alcon, Fort Worth, TX) was injected intracamerally in the right eye of another 5 rabbits as a control group. A slit lamp examination, IOP measurement, specular microscope, and pachymetry were performed at before injection and at 1, 2, and 4 weeks after the intracameral injection of bevacizumab and BSS. The IOP was measured by using a Tonopen-XL (Mentor, Co.) and the cornea thickness was measured by ultrasonic pachymetry (Model P55, Paradigm Medical Industries, Inc.) after the topical application of proparacaine HCl 0.5%. All rabbit eyes were photographed through a specular microscope (Topcon SP-2000, Japan) before and after the intracameral injections. Images from the center of the cornea of each eye were obtained. After 20 endothelial cells in each image had been manually designated, the cell density was calculated using the instrument's computer program. Comparisons of the endothelial cell counts, corneal thickness, and IOP were tested with the Mann-Whitney test and Wilcoxon signed test among and between each group. P values of less than 0.05 were considered to be statistically significant.

One rabbit from each group was euthanized with a fatal dose of pentobarbital sodium at 1 week and 1 month after the intracameral injections. After enucleation, the corneal button was excised atraumatically. For the scanning electron microscopic examination, the specimens were fixed in 2.5% glutaraldehyde for 2 hours, washed in phosphate buffer (pH 7.4), postfixed in 2% osmium tetroxide in phosphate buffer (pH 7.4), and dehydrated in increasing concentrations of ethanol. The specimens were critical-point dried, mounted on metal stubs with a conductive silver paint, and then sputtered with a 10 nm thick layer of gold using an E-102 ion sputter (Hitachi, Japan). The tissue samples were examined on an H-2500 scanning electron microscope (Hitachi, Japan) at an acceleration voltage of 25 kV.

Back to Top | Article Outline
Effect in NVG Combined With AGV Implantation

A retrospective review of the medical records of 6 patients who received AGV implant surgery and an intracameral bevacizumab injection was carried out on patients with NVG unresponsive to the maximal medical treatment. Informed consent was obtained from all patients after gaining approval for the off-label use of bevacizumab from the institutional review board committee, and discussing the experimental nature of the treatment. The surgical procedure of the AGV implantation was as follows.

A fornix-based conjunctival flap and dissection of the tenon's capsule were performed at the superotemporal quadrant to place the plate of the implant into the subtenon's space. Before inserting the AGV, the implant was primed by irrigating the lumen of the tube with a balanced salt solution. The plate was fixed to the sclera with two 8-0 black nylon sutures 8 mm behind the limbus. The tube was cut to the desired length, with its sharp bevel facing anteriorly. An anterior chamber (AC) paracentesis wound was created at the peripheral cornea and 1% sodium hyaluronate (Healon, Pharmacia and Upjohn) was injected into the AC to prevent sudden collapse of the AC. The implant tube was inserted into the AC, parallel to the iris plane, through the sclerostomy that had been made with a 23-gauge syringe needle. The tube was fixed to the sclera with 9-0 black nylon sutures. The anterior part of the silicone tube was covered with a donor scleral graft (4×4 mm) and fixed to the sclera with 10-0 black nylon sutures. The conjunctiva was closed with 8-0 vicryl sutures. A intracameral bevacizumab injection (1.25 mg/0.5 mL) was performed at the end of the surgical procedures. No adjunctive antimetabolite was used in any of the patients. Topical 0.3% levofloxacin (Cravit, Santen, Osaka, Japan) and 1% prednisolone acetate eye drops (Pred Forte, Allergan, Irvine, CA) were administered 4 times daily. Ocular examinations, including the best corrected visual acuity using the Snellen chart, IOP measurements using the Goldmann applanation tonometer, slit lamp examination, and also gonioscopic and fundus examinations, were performed preoperatively and postoperatively on day 1 and at 1, 4, and 8 weeks, respectively. The visual acuity, such as counting fingers, hand motion, and light perception, was converted into the Snellen acuity of 20/2400, 20/4800, and 20/9600, respectively.21 Iris fluorescein angiography (IFA) was performed preoperatively and at 1 to 2 weeks after surgery to evaluate the changes in INV. All ocular and systemic complications, such as uncontrolled hypertension or thromboembolic events after the intracameral bevacizumab injection and AGV implantation surgery, were recorded.

Back to Top | Article Outline


Corneal Toxicity in Rabbit Eyes

The corneal endothelial cell count and central corneal thickness were not significantly different in the intracameral bevacizumab and intracameral BSS injection groups. The IOP did not change significantly in the bevacizumab and BSS injection groups (Table 1).

Table 1
Table 1
Image Tools

The scanning electron microscopic images showed a normal corneal endothelial cell appearance in both the bevacizumab and BSS injection groups after 1 month (Fig. 1). The hexagonal shape of the corneal endothelium was well preserved. No abnormal endothelial cells, such as bleb formation and disintegration of cellular border, were observed. Smooth and distinct cell borders were detected between each endothelial cell. The intercellular border thickness also appeared normal in both groups.

Figure 1
Figure 1
Image Tools
Back to Top | Article Outline
Effect in NVG Combined With AGV Implantation

Six patients (3 male and 3 female) were enrolled in this study. The causes of NVG were central retinal vein occlusion (2 patients), ocular ischemic syndrome (1 patient), and proliferative diabetic retinopathy (PDR, 3 patients). The mean age of the 6 patients was 66.3±14.2 years (range, 45 to 83 y). The mean IOP before surgery was 44.2±9.4 mm Hg under the maximal tolerated topical and systemic medication. The visual acuity ranged from light perception to counting fingers. None of the 6 patients had received any previous intracameral or intravitreal anti-VEGF treatments.

Regression of the INV was observed within 7 days (as early as 2 d) in all patients after the intracameral injection of bevacizumab and AGV implantation. The iris fluorescein angiogram showed regressed INV in all patients. In case 1, the neovascularization of angle (NVA) was also regressed on the gonioscopic examination. There were no complications, such as a flat AC, hyphema, and hypotony, in the postoperative periods in all 6 cases. Two cases (cases 2 and 3) showed a hypertensive phase (IOP >21 mm Hg during follow-up) at 2 months in case 2 and at 6 months in case 3, but returned to normal IOP levels (≤21 mm Hg) with topical antiglaucoma medications. No systemic complications associated with intracameral bevacizumab injection were recorded. Table 2 provides a summary of the 6 cases. The following gives representative examples of the clinical cases according to their underlying retinal pathology.

Table 2
Table 2
Image Tools
Back to Top | Article Outline
Case 1

A 55-year-old man, who had previously been treated for open angle glaucoma in both eyes, visited our clinic complaining of ocular pain in his left eye over the previous 2 weeks. The visual acuity was counting fingers, which was converted to a corresponding Snellen acuity of 20/2400 for comparison purposes,21 and the IOP was 45 mm Hg at presentation. Iris and angle rubeosis were observed on the gonioscopic examination (Fig. 2A). The fundoscopic examination and fluorescein angiography revealed central retinal vein occlusion in the left eye. Prompt panretinal photocoagulation (PRP) was performed and AGV implantation with an intracameral bevacizumab injection was carried out 2 days after PRP. IFA performed before surgery showed leaking pathologic vessels around the pupillary margin (Fig. 2B). Five days after surgery, the slit lamp and gonioscopic examinations revealed regression of the INV and NVA (Fig. 2C). Five days after the intracameral bevacizumab injection, the IFA showed no leakage in the pupillary margin at the early phase but slightly leaking vessels at the late phase (Fig. 2D). Four weeks after surgery, the patient's visual acuity recovered to 4/200 and the IOP was 13 mm Hg without antiglaucoma medication. There was no recurrence of the INV and NVA during the follow-up periods.

Figure 2
Figure 2
Image Tools
Back to Top | Article Outline
Case 3

A 78-year-old female patient presented with a visual disturbance and ocular pain in her right eye with 3 months duration. At presentation, her visual acuity was hand motions (equivalent to Snellen acuity of 20/4800)21 and the IOP was 46 mm Hg. An anterior segment examination revealed INV around the pupillary margin (Fig. 3A), and the fundus examination revealed severe attenuated arterioles and also pale and total cupping of the optic disc. IFA showed leaking, abnormal vessels on the iris surface (Fig. 3B). More than 80% stenosis of the internal carotid artery was observed in the carotid Doppler echography. Two days after AGV implantation with the intracameral bevacizumab injection, the INV had almost completely regressed in the slit lamp examination (Fig. 3C). The IFA showed no leakage in the early phase but slight leakage in the late phase (Fig. 3D). At 4 weeks after surgery, the patient's visual acuity was still hand motions (equivalent to Snellen acuity of 20/4800)21 and the IOP was 21 mm Hg without antiglaucoma medication. There was no recurrence of INV during the follow-up examinations.

Figure 3
Figure 3
Image Tools
Back to Top | Article Outline
Case 6

A 45-year-old diabetic patient, who had pars plana vitrectomy for PDR 2 months earlier, visited the clinic complaining of ocular pain and decreased vision in his left eye over the previous 3 days. His visual acuity was counting fingers (equivalent to Snellen acuity of 20/2400)21 and the IOP was 35 mm Hg. The slit lamp examination revealed INV and mild hyphema. AGV implantation and an intracameral bevacizumab injection were performed. Five days after surgery, the INV regressed and the hyphema had disappeared. However, the INV reappeared after 3 months despite the well-controlled IOP (Fig. 4A). Therefore, a second injection of intracameral bevacizumab was performed. The IFA before the second injection showed abnormal vessels on the iris surface (Fig. 4B). After 7 days, a slit lamp examination showed regression of the INV (Fig. 4C), and the IFA showed reduced leakage from abnormal vessels (Fig. 4D). The patient's visual acuity recovered to 4/16 and the IOP was 18 mm Hg 4 weeks after the second intracameral bevacizumab injection. However, the INV recurred 8 weeks after the second intracameral bevacizumab injection.

Figure 4
Figure 4
Image Tools
Back to Top | Article Outline


Most studies have reported that an intravitreal bevacizumab injection is not toxic to the retina.19,20 However, little is known about the corneal toxicity of bevacizumab after an intracameral or intravitreal injection. Yoeruek et al22 reported that bevacizumab was not toxic to cultured corneal cells, and Kernt et al23 also reported no cellular toxicity of bevacizumab in commonly used concentrations (1 to 1.25 mg) in an in vitro model. In this study, there was no corneal toxicity after the intracameral injection of bevacizumab (1.25 mg/0.05 mL) for a period of 1 month in rabbit eyes. A total of 1.25 mg of bevacizumab was used in this study because it is a common intravitreal dosage and the AC volume of a rabbit is too small to inject more than 0.05 mL (1.25 mg) of bevacizumab without causing an increase in IOP, which can affect the corneal endothelium and cause corneal edema.

Although Grisanti et al24 reported the rapid regression of iris rubeosis after an intracameral bevacizumab injection, it is not popular to inject bevacizumab intracamerally. Most studies reported similar results after an intravitreal injection.9–17 This study is the first report of an intracameral bevacizumab injection in conjunction with AGV implantation.

The pharmacokinetics of bevacizumab after an intracameral injection is not known. Most drugs cleared faster from the eye when administered intracamerally than intravitreally. Intracameral bevacizumab might clear faster from the AC than after an intravitreal injection. However, in this study and that of Grisanti et al,24 the effect of an intracameral bevacizumab injection in the regression of iris and angle neovascularization was comparable with a previous report of an intravitreal injection.9–17 It is possible that an intracameral injection would be an alternate method of administering bevacizumab in neovagular glaucoma. In this study, viscoelastics were injected into the AC to prevent the sudden collapse of the AC. Viscoelastics may act as a drug delivery system and interfere with the clearance of bevacizumab from the AC.25

Some studies reported that INV recurred 2 to 3 months after the intravitreal bevacizumab injection.9,10 In the present cases, 3 patients with PDR experienced recurring INV 2 to 3 months after the intracameral bevacizumab injection and AGV implantation. In the PDR cases, the regression of INV was incomplete and not as dramatic as in the other cases. Therefore, it was assumed that there was some difference in the response to bevacizumab according to the underlying disease process.

Some reports showed that a single intravitreal bevacizumab injection was enough to control the IOP in NVG without the need for any further surgical procedures.9,12,14,16 However, in many cases, the IOP reduction after the intravitreal bevacizumab injection was not as dramatic as the regression of INV and NVA. Therefore, additional surgical procedures, such as drainage device surgery or cyclodestructive procedure, were required to control the IOP.9,11–13 In the present cases, AGV implantation and intracameral bevacizumab injections were performed with the expectation of rapid IOP control and the regression of INV in the early postoperative period. The IOP was controlled successfully and INV had regressed dramatically within 1 week. It is reasonable to perform a combined procedure if a patient has visual potential in NVG.

Yazdani et al9 reported a case series of stable IOP after an intravitreal bevacizumab injection even in a total synechial angle closure of NVG. However, Gheith et al17 reported that patients with peripheral anterior synechia required subsequent glaucoma surgery. In the present cases, none of the patients presented with total synechial angle closure. Additional studies will be needed to determine if bevacizumab alone can cause INV/NVA regression and control the IOP without the need for other surgical procedures in total synechial angle closure.

These results suggest that intracameral bevacizumab is a safe and effective treatment for INV/NVA when combined with AGV implantation, and may be considered as an adjuvant for other treatments in patients with NVG, such as PRP or glaucoma tube shunt procedures. The limitation of this study is that it presents a small number of cases without a control group, and the systemic safety regarding intracameral bevacizumab injections were not adequately measured. Therefore, randomized case-control trials will be needed to further evaluate the effectiveness, effect duration, and need for repeated injections.

Back to Top | Article Outline


1. Ferrara N. Vascular endothelial growth factor: basic science in clinical progress. Endocrinol Rev. 2004;25:581–611.

2. Tripathi RC, Li J, Tripathi BJ, et al. Increased level of vascular endothelial growth factor in aqueous humour of patients with neovascular glaucoma. Ophthalmology. 1998;105:232–237.

3. Aiello LP, Avery RL, Arrigg PG, et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med. 1994;331:1480–1487.

4. Adamis AP, Miller JW, Bernal MT, et al. Increased vascular endothelial growth factor levels in the vitreous of eyes with proliferative diabetic retinopathy. Am J Ophthalmol. 1994;118:445–450.

5. Sivak-Callcott JA, O'Day DM, Gass DM, et al. Evidence-based recommendations for the diagnosis and treatment of neovascular glaucoma. Ophthalmology. 2001;108:1767–1776.

6. Lai JS, Poon AS, Chua JK, et al. Efficacy and safety of the Ahmed glaucoma valve implant in Chinese eyes with complicated glaucoma. Br J Ophthalmol. 2000;84:718–721.

7. Topouzis F, Coleman AL, Choplin N, et al. Follow-up of the original cohort with the Ahmed glaucoma valve implant. Am J Ophthalmol. 1999;128:198–204.

8. Huang MC, Netland PA, Coleman AL, et al. Intermediate-term clinical experience with the Ahmed Glaucoma Valve implant. Am J Ophthalmol. 1999;127:27–33.

9. Yazdani S, Hendi K, Pakravan M. Intravitreal bevacizumab (Avastin) injection for neovascular glaucoma. J Glaucoma. 2007;16:437–439.

10. Oshima Y, Sakaguchi H, Gomi F, et al. Regression of iris neovascularization after intravitreal injection of bevacizumab in patients with proliferative diabetic retinopathy. Am J Ophthalmol. 2006;142:155–158.

11. Davidorf FH, Mouser JG, Derick RJ. Rapid improvement of rubeosis iridis from a single bevacizumab (Avastin) injection. Retina. 2006;26:354–356.

12. Mason JO III, Albert MA Jr, Mays A, et al. Regression of neovascular iris vessels by intravitreal injection of bevacizumab. Retina. 2006;26:839–841.

13. Iliev ME, Domig D, Wolf-Schnurrbursch U, et al. Intravitreal bevacizumab (Avastin) in the treatment of neovascular glaucoma. Am J Ophthalmol. 2006;142:1054–1056.

14. Batioğlu F, Astam N, Özmert E. Rapid improvement of retinal and iris neovascularization after a single intravitreal bevacizumab injection in a patient with central retinal vein occlusion and neovascular glaucoma. Int Ophthalmol. 2008;28:59–61.

15. Vatavuk Z, Bencic G, Mandic Z. Intravitreal bevacizumab for neovascular glaucoma following central retinal artery occlusion. Eur J Ophthalmol. 2007;17:269–271.

16. Silva Paula J, Jorge R, Alves Costa R, et al. Short-term results of intravitreal bevacizumab (Avastin) on anterior segment neovascularization in neovascular glaucoma. Acta Ophthalmol Scand. 2006;84:556–557.

17. Gheith ME, Siam GA, de Barros DS, et al. Role of intravitreal bevacizumab in neovascular glaucoma. J Ocul Pharmacol Ther. 2007;23:487–491.

18. Barki SJ, Snyder MR, Reid JM, et al. Pharmacokinetics of intravitreal bevacizumab (Avastin). Ophthalmology. 2007;114:855–859.

19. Feiner L, Barr EE, Shui YB, et al. Safety of intravitreal injection of bevacizumab in rabbit eyes. Retina. 2006;26:882–888.

20. Shahar J, Avery RL, Heilweil G, et al. Electrophysiologic and retinal penetration studies following intravitreal injection of bevacizumab (Avastin). Retina. 2006;26:262–269.

21. Edwards A, Fishman GA, Anderson RJ, et al. Visual acuity and visual field impairment in Usher syndrome. Arch Ophthalmol. 1998;116:165–168.

22. Yoeruek E, Spitzer MS, Tatar O, et al. Safety profile of bevacizumab on cultured human corneal cells. Cornea. 2007;26:977–982.

23. Kernt M, Welge-Lüssen U, Yu A, et al. Bevacizumab is not toxic to human anterior- and posterior-segment cultured cells. Ophthalmologe. 2007;104:965–971.

24. Grisanti S, Biester S, Peters S, et al. Intracameral bevacizumab for iris rubeosis. Am J Ophthalmol. 2006;142:158–160.

25. Liao YH, Jones SA, Forbes B, et al. Hyaluronan: pharmaceutical characterization and drug delivery. Drug Deliv. 2005;12:327–342.


Ahmed valve; bevacizumab; intracameral injection; neovascular glaucoma

© 2009 Lippincott Williams & Wilkins, Inc.


Article Level Metrics

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.