We performed a study of a consecutive series of 54 eyes of 50 patients who were diagnosed with NVG in Tianjin Eye Hospital from May 2015 to March 2017. Fifty-two eyes of 48 patients received intravitreal conbercept combined with trabeculectomy, the other 2 eyes had absolute glaucoma with unbearable pain, and enucleation was performed to identify iris neovascularization by light microscopy as a control.
The inclusion criteria were (1) NVG patients whose underlying diseases were DR, CRVO, carotid artery stenosis or occlusive disease, (2) patients without intravitreal anti-VEGF agents or PRP before, (3) patients who were 18 to 85 years old, (4) patients with a visual acuity ≥ hand motion, (5) patients with an IOP >21 mmHg (1mmHg=0.133kPa) and under the maximum dose of antiglaucoma medications, (6) patients with iris or angle neovascularization, and (7) patients with or without pupillary margin pigmented layer eversion. The exclusion criteria were (1) NVG combined with ocular tumours or uveitis, (2) severe cardiovascular and cerebrovascular diseases, (3) patients who had recent active inflammation of the eye, or (4) patients who failed to complete the scheduled follow-ups for various reasons. The study was carried out with the approval of the ethics committee of Tianjin Eye Hospital (No. TJYYLL-2015-15), and all patients provided voluntary informed consent to participate in the study.
Preoperative evaluation included the best corrected visual acuity (BCVA), slit-lamp microscopy examination, gonioscopy examination, IOP, primary disease classification and other measures. BCVA was converted to the logarithm of the minimum angle of resolution (LogMAR), IOP was measured with a Goldmann tonometer (AT 900, Haag-Strsit AG, Switzerland, Bern), but if patients had corneal oedema due to high IOP, an Icare handheld rebound tonometer (TA01i, Finland Aike company, Finland, Helsinki) was used. Iris neovascularization grading was as follows: grade 1 was surface neovascularization of the pupillary zone of the iris involving ≤2 quadrants, grade 2 was surface neovascularization of the pupillary zone of the iris involving ≥2 quadrants, grade 3 was neovascularization of the ciliary zone of the iris involving 1 to 3 quadrants addition to the pupillary zone, and grade 4 was neovascularization of the ciliary zone of the iris involving ≥3 quadrants .
All patients were assessed by one glaucoma specialist. Under sterilization and topical anaesthesia (0.5% proparacaine hydrochloride eye drops), anterior chamber paracentesis was performed in 48 patients (52 eyes) with high IOP, and then 0.05 ml (0.5mg) conbercept was injected intravitreally via the pars plana, approximately 3.5 mm from the limbus, with a 30G needle. Then, IOP and light perception were examined. Patients were given topical antibiotics and antiglaucoma medications after surgery.
Trabeculectomy was performed when the neovascularization of the iris surface regressed and anterior chamber inflammation was relieved (the interval time was 2 to 7 days). Under peribulbar anaesthesia, a fornix-based conjunctival flap and a half thickness of 4 mm*3 mm square scleral flap was made, a mitomycinC (MMC)(0.4 mg/ml) or 5-fluorouracil (5-FU)(25 mg/ml)-soaked sponge was placed in the scleral flap for 2-3 min before rinsing thoroughly with 30 ml saline. Trabecular meshwork (2 mm*1 mm) was cut, and the peripheral iridectomy was performed. The scleral flap was closed with two 10-0 nylon sutures at its corners, and the conjunctiva was sutured with 10-0 nylon sutures.
Two patients (2 eyes) had enucleation under retrobulbar anaesthesia, the bulbar conjunctiva was cut along the limbus corneae, and then the conjunctiva was separated, the tenon capsule and sclera were bluntly dissected to the equator, the extraocular muscles were cut, the optic nerve and soft tissue were removed, the eye was removed, a enucleated eye was implanted in the tenon capsule, the front fascia was closed by pouch suture, the tenon capsule was packed with Vaseline gauze, and the eye socket was placed.
Three weeks or 1 month later, PRP was performed (577 nm yellow light supra scan laser light, Quantel, France, Clermont-Ferrand), which was completed 2-3 times with 250-350 mv energy (appear II-III spot as the standard), an interval time of 0.03 s, a spot diameter of 200 μm, and a total light solidifying point of 1500-2000 points.
The surgically excised trabecular specimens of 26 eyes were immediately immersed in a solution of 10% formalin fixative at room temperature for 24 hours, and then these specimens were dehydrated with different concentrations of alcohol (from 70% to 100%) and made transparent with dimethylbenzene two times. After embedding in paraffin, 5-μm-thick sections were made and allowed to dry overnight. Then, these specimens were stained with haematoxylin-eosin and observed with light microscopy (Leica 400B, Leica company, Germany, Solms).
The iris tissue of the other eyes were immersed in 4% glutaraldehyde and 1% osmic acid solution at 4°C for 4 hours and 1 hour, then these specimens were washed with PBS powder and were dehydrated with gradient acetone. After embedding in 812 resin, these specimens were cut into 60 nm sections and stained with lead citrate and uranyl acetate. Then, transmission electron microscopy (JEM-1230, JEOL company, Japan, Tokyo) was used to observe these sections.
Then, 0.1-0.15 ml aqueous humour samples were collected during intravitreal injection and trabeculectomy and were placed into a sterile Eppendorf tube and rapidly frozen at -80°C for use (all samples were obtained at the beginning of surgery to avoid breakdown of the blood-aqueous barrier that is associated with surgical trauma). The main reagents were human VEGF-A enzyme-linked immunoabsorbent assay (ELISA) kit, human TGF- β1 ELISA kit and human PLGF ELISA kit (China BlueGene Co., China, Shanghai). The concentrations of VEGF-A, TGF- β1 and PLGF in aqueous humour were determined by the double antibody sandwich ELISA method. The standard curve for ELISA (use a four parameter logistic curve fit) was set up, the desired numbers of coated wells were secured in the holder, and then 100 μL of standards or samples were added to the appropriate well, adding 100μL of PBS (pH 7.0-7.2) to the blank control well, adding 50 μL of conjugate to each well (not the blank control well), and covering and incubating the plate for 1 hour at 37°C . The plate was washed five times with diluted wash solution (350-400 μL/well/wash) using an auto washer, and then the plate was dried. Next, 50 μL substrate A and 50 μL substrate B were added to each well, covering and incubating for 15 minutes at 37°C, then adding 50 μL of stop solution to each well, immediately determining the optical density (O.D.) at 450 nm using a microplate reader (ST-360, Shanghai KEHUA Experimental System Co., China, Shanghai).
The postoperative follow-up was 1 year. Patients were followed on schedule (post-injection and 1 day, 1 week, 1 month, 3 months, 6 months, and 1 year after trabeculectomy). BCVA and IOP were recorded. Slit-lamp microscopy examination, gonioscopy examination, the numbers of anti-glaucoma medications, and intraoperative and postoperative complications were recorded. Previous study shown that complete success of the surgery was defined as IOP ≤21 mmHg without any topical ocular hypotensive medication, partial success was defined as IOP ≤21 mmHg with topical ocular hypotensive medication , and surgical failure was defined as IOP >21 mmHg at 2 consecutive follow-up visits even with anti-glaucoma medication or additional glaucoma surgeries, such as filtration surgery and cyclophotocoagulation, or loss of light perception . But 21mmHg of IOP for these eyes is still very high, so the cue off should be 18mmHg, and complete and partial success based on cut off the IOP of 21, 18 and 15 mmHg at various time points will be estimated to give a detailed overview of the exact results.
Statistical analysis was conducted using SPSS 23.0 software (SPSS Inc., America, Chicago). These data on IOP, BCVA and the concentrations of cytokines were confirmed by W test, which was consistent with a normal distribution. Therefore, these results were presented as the mean±standard deviation (SD). The data on the numbers of antiglaucoma drugs were not consistent with a normal distribution. Therefore, the results were expressed as M (Q1, Q3). Single effect repeated-measure analysis of variance and Dunnett’s t-test were used to assess differences between BCVA and IOP at different time points. The levels of VEGF-A, TGF-β1 and PLGF were compared by one-way ANOVA and LSD t-tests. The numbers of antiglaucoma medications were compared by K-W tests of multiple groups of independent samples. The correlation analysis between iris neovascularization and VEGF-A, TGF-β1, PLGF were completed by Spearman rank correlation coefficients. Kaplan-meier survival analysis was performed to estimate the surgical success. A P values of less than 0.05 were considered statistically significant.