Purpose
To examine the histopathologic effect of a single intravitreal injection of bevacizumab on newly formed vessels in eyes with proliferative diabetic retinopathy (PDR).
Design
Interventional case series and laboratory investigation.
Methods
Two days after intravitreal injection of bevacizumab (1.25 mg/eye), pars plana vitrectomy or trabeculectomy was performed for the treatment of PDR or neovascular glaucoma (NVG) associated with PDR. Ten surgically removed preretinal proliferative tissues and 6 deep scleral flaps containing trabecular meshwork were fixed in 2% glutaraldehyde or 4% paraformaldehyde and were subjected to transmission electron microscopic analysis, immunohistochemical analysis, and terminal deoxyuridiine triphosphate (dUTP) nick-end labeling staining. Two surgically removed preretinal proliferative tissues and 2 deep scleral flaps from patients with PDR and NVG, but without preoperative intravitreal injection of bevacizumab (IVB), served as controls.
Results
In control tissues, vascular endothelial cells possessed many fenestrations and were accompanied by pericytes. Apoptotic vascular endothelial cells frequently were observed in tissue after intravitreal injection of bevacizumab, whereas they were not observed in control tissues. Additionally, no apparent fenestration was observed in newly formed vessels from either proliferative tissue or trabecular meshwork after intravitreal injection of bevacizumab. In both PDR and NVG tissues after intravitreal injection of bevacizumab, overexpression of smooth muscle actin was observed in newly formed vessels, suggesting that the treatment may have increased pericytes on the vasculature as compared with control tissue.
Conclusions
Intravitreal injection of bevacizumab may induce changes in immature, newly formed vessels of PDR or NVG tissue, leading to endothelial apoptosis with vascular regression, while inducing normalization of premature vessels by increasing pericyte coverage and reducing vessel fenestration.
Proliferative Diabetic Retinopathy (PDR) can cause severe visual loss resulting from complications such as vitreous hemorrhage, tractional retinal detachment, and development of neovascular glaucoma (NVG). Vascular endothelial growth factor (VEGF) is known to play a central role in ocular neovascularization. Panretinal photocoagulation remains the first-line therapy for ocular neovascularization. One of the mechanisms underlying the benefit of panretinal photocoagulation now is thought to involve inhibition of VEGF production, thus inducing regression of neovascularization.
Bevacizumab (Avastin; Roche, Reinach, Switzerland) is a recombinant humanized monoclonal anti-VEGF antibody first used for the treatment of metastatic colorectal cancer. However, many reports of off-label therapy in ophthalmology have been published in recent years. In these reports, bevacizumab has been reported to lead to the regression of neovascularization in eyes with age-related macular degeneration, PDR, and NVG, as well as reducing macular edema associated with diabetic retinopathy and retinal vein occlusion. However, the histopathologic effects of intravitreal injection of bevacizumab are not fully understood, especially in human eyes. The present study investigated the histopathologic changes in the neovascular tissues of diabetic patients with PDR and NVG after intravitreal injection of bevacizumab.
Methods
Sixteen neovascular tissue specimens removed from 15 patients with PDR and NVG were examined by histopathologic analysis from April 2008 through October 2008 at Kyushu University Hospital. During this time, other diabetic vitrectomies were performed in our hospital without intravitreal injection of bevacizumab; these cases were not believed to be at high risk of intraoperative bleeding or postoperative rubeosis. The clinical data of the patients are summarized in the Table . All patients had neovascularization of the disc and neovascularization elsewhere noted along the arcade vessels. Inclusion criteria for the study were: tractional retinal detachment with active neovascularization within the preretinal proliferative membrane, repeated vitreous hemorrhage with active neovascularization, rubeosis with vitreous hemorrhage precluding additional panretinal photocoagulation, and refractory NVG.
| Patient or Control No. | Diagnosis | Age (yrs) | Duration of Diabetes (yrs) | GHbA1c (%) | Previous Therapies | Systemic Disorder |
|---|---|---|---|---|---|---|
| Patients | ||||||
| 1 | PDR | 40 | 15 | 12.4 | LP | DM |
| 2 | PDR | 42 | 13 | 13.1 | LP | DM, AH |
| 3 | PDR | 48 | 2 | 12.4 | LP | DM, AH |
| 4 | PDR | 30 | 8 | 11.2 | LP | DM |
| 5 | PDR | 30 | 8 | 11.2 | LP | DM |
| 6 | PDR | 51 | 30 | 8.4 | LP | DM |
| 7 | PDR | 26 | 3 | 10.6 | LP | DM |
| 8 | PDR | 71 | 20 | 6.8 | LP | DM |
| 9 | PDR | 49 | 20 | 7.9 | LP | DM |
| 10 | PDR | 42 | 6 | 8.4 | LP | DM, AH |
| 11 | NVG | 63 | 35 | 6.5 | LP, PPV | DM, AH |
| 12 | NVG | 50 | 27 | 7.9 | LP, PPV | DM, AH |
| 13 | NVG | 50 | 27 | 7.9 | LP, PPV | DM, AH |
| 14 | NVG | 72 | 10 | 11.0 | LP, PPV | DM |
| 15 | NVG | 53 | 8 | 10.6 | LP, PPV | DM, AH |
| 16 | NVG | 59 | 20 | 9.1 | LP, PPV | DM |
| Controls | ||||||
| 1 | PDR | 57 | 30 | 9.6 | LP | DM, AH |
| 2 | PDR | 56 | 19 | 9.9 | LP | DM, AH |
| 3 | NVG | 58 | 8 | 7.4 | LP, PPV | DM, AH |
| 4 | NVG | 40 | 11 | 11.6 | LP, PPV | DM |
In cases of uncontrolled hypertension, or in cases with a history of thromboembolic events including myocardial infarction, and in cases of cerebral insult or renal disease, no off-label use of intravitreal bevacizumab was offered. Intravitreal injection of bevacizumab was performed 2 days before surgery. During the vitrectomy, phacoemulsification and intraocular lens implantation were performed simultaneously if needed. During trabeculectomy, a deep scleral flap was removed after creating a scleral flap. Deep scleral flaps and surgically removed membranes containing neovascularization of the disc were immediately fixed in 4% paraformaldehyde or 2% glutaraldehyde at 4 C for 24 hours and were subjected to histopathologic analysis.
Intravitreal Injection of Bevacizumab
Treated eyes received 1 single intravitreal injection of 1.25 mg (0.05 mL) bevacizumab (25 mg/mL) via the pars plana 4 mm posterior to the limbus using a 30-gauge needle. When removing the needle, the injection site was compressed with forceps. Two days after intravitreal injection of bevacizumab, pars plana vitrectomy or trabeculectomy was performed.
Immunohistochemical Analysis
Surgically removed membranes and deep scleral flaps were fixed in 4% paraformaldehyde, embedded in paraffin wax, cut into 5-μm sections, and deparaffinized according to standard procedures: anti-VEGF (dilution, 1:100, rabbit; Santa Cruz Biotechnology, Santa Cruz, California, USA), anti-CD34 (dilution, 1:100; Novocastra, Newcastle, United Kingdom), and anti–smooth muscle cell actin (dilution, 1:100, mouse immunoglobulin G; DAKO, Glostrup, Denmark). Heat-induced epitope retrieval was performed by immersing sections of tissue in citrate buffer (pH 6.0). The surgically removed deep scleral flaps obtained by a trabeculectomy contained the Schlemm canal, juxtacanalicular connective tissue, and almost all parts of the corneoscleral meshwork.
Terminal Deoxyuridiine Triphosphate (dUTP) Nick-End Labeling Staining
The terminal dUTP nick-end labeling (TUNEL) procedure and quantification of TUNEL-positive cells were performed using an Apoptosis Detection TACS TdT Kit (R&D Systems, Minneapolis, Minnesota, USA) for paraffin sections according to the instructions of the manufacturer. The number of TUNEL-positive cells was counted in a masked manner.
Transmission Electron Microscopic Analysis
Surgically removed membranes and deep scleral flaps were prefixed with 2% glutaraldehyde in 0. 1 M CacoB. After rinsing, samples were postfixed for 90 minutes with 1% osmium tetroxide in 0.1 M CacoB, dehydrated through a graded ethanol series, stained en bloc during dehydration with a saturated solution of uranyl acetate in 70% ethanol, and embedded in Epon (EM Japan, Tokyo, Japan). Ultra-thin (60-nm) sections were cut from blocks and were mounted on copper grids. The sections were stained with lead citrate and uranyl acetate and were examined under a H-7650 electron microscope (Hitachi, Tokyo, Japan).
Results
Immunohistochemical Analysis of Proliferative Membranes and Trabecular Meshwork
In control eyes not receiving preoperative intravitreal injection of bevacizumab, PDR membranes and NVG trabecular meshwork contained many immature vessels that were CD34 immunopositive and smooth muscle actin immunonegative, suggesting vessels without significant pericyte coverage. Mature vessels with a CD34/smooth muscle actin double-immunopositive characteristic were observed less often in these control tissues ( Figure 1 ). In contrast, almost all vessels within PDR membranes and NVG trabecular meshwork from eyes receiving intravitreal injection of bevacizumab were immunopositive both for CD34 and smooth muscle actin. Conversely, vessels with a small capillary-like structure that were CD34 immunopositive and smooth muscle actin immunonegative decreased in number ( Figure 1 ).
