Anterior Retinal Cryotherapy and Intravitreal Injection of Bevacizumab in the Treatment of Nonclearing Vitreous Hemorrhage in Proliferative Diabetic Retinopathy

Abstract

Purpose:

To evaluate the efficacy of intravitreal bevacizumab (IVB) injection combined with anterior retinal cryotherapy (ARC) in the treatment of proliferate diabetic retinopathy (PDR) with vitreous hemorrhage (VH).

Methods:

A retrospective case controlled study was performed on 67 cases (67 patients) with PDR and persistent VH obscuring the fundus detail for ≧3 weeks. The follow-up period was ≧6 months. Cases with extensive vitreo-retinal adhesion by ultrasonography were excluded. In the study group, transconjunctival ARC (a total of 12 spots) followed by IVB 1.25 mg was performed in the same setting; in the control group, only IVB was given. Panretinal photocoagulation was performed when possible. Repeated IVB was performed in both groups if VH persisted for >4 weeks following the initial treatment. Vitrectomy was performed if VH showed no improvement for 12 to 16 weeks. The vitreous clear-up time (VCUT) in the posterior pole and the rate of vitrectomy were compared between the 2 groups.

Results:

In the study group (n=35), second treatment was required in 7 out of 35 patients (20.0%). Vitrectomy was performed in 5 (14.3%) patients. In the control group (n=32), 8 out of 32 eyes (25.0%) underwent second injection. Four eyes (12.5%) had vitrectomy within the follow-up time. VCUT in the study and the control group was 6.9±8.7 weeks and 13.0±9.3 weeks respectively (P=0.003).

Conclusion:

Combined treatment of ARC and IVB is associated with more rapid clearing of VH in eyes with PDR compared with IVB alone.

Introduction

Vitreous hemorrhage (VH) has been one of the leading causes of compromised vision in eyes with proliferative diabetic retinopathy (PDR).¹ It not only precludes timely panretinal photocoagulation (PRP) but also stimulates fibrovascular proliferation (FVP), forming a vicious circle. Conventionally, persistent VH was treated with indirect panretinal laser photocoagulation if the media is clear enough.^2–5 Peripheral retinal cryotherapy has become another treatment option for PDR with dense VH.^6–10 However, both treatments are limited by the less than satisfactory response of VH reabsorption.

Recently, bevacizumab (Avastin) has been shown to induce rapid regression of retinal and iris neovascularization.^11–13 During the clearing process following acute VH, bevacizumab may prevent the occurrence of multiple episodes of rebleeding that prolong the vitreous clear-up time (VCUT).¹⁴ Our previous study has demonstrated rapid vitreous clear-up may be achieved by intravitreal bevacizumab (IVB). If combined with subsequent PRP, the need for surgical intervention may be significantly reduced.¹⁵ Although the duration of VH can be reduced by IVB, more than half of the patients still suffered from prolonged (≧10 weeks) vision decrease.¹⁵ Peripheral cryotherapy has a similar effect as PRP; in addition, it may facilitate blood reabsorption through mechanism different from that with bevacizumab treatment.^16–18 We hypothesized that the combination treatment of anterior retinal cryotherapy (ARC) and IVB may further facilitate blood reabsorption, suppress neovascularization, and achieve more rapid resolution of VH in PDR. To confirm this hypothesis, a retrospective study was conducted to investigate the efficacy of ARC with IVB in treating PDR complicated by persistent VH.

Methods

From September 2008 to March 2010, consecutive patients suffering from PDR with persistent VH treated with ARC and IVB or IVB alone were selected for the study. Ophthalmic examinations were performed by a single retinal specialist (C.M.Y.), with results confirmed by another (C.H.Y.). Two ophthalmologists (C.M.Y. and C.H.Y.) performed the treatment. The potential pros and cons of both surgical procedures were well explained to every patient before operation. In short, we specifically informed the patients that our previous study had shown IVB might shortened the vitreous clear-up time;¹⁹ the addition of cryotherapy might need peribulbar anesthesia, induce temporal tissue swelling, and have the potential of inducing more traction. We also stated that, according to our clinical experience, the addition of cryotherapy might prompt VH reabsorption more rapidly than bevacizumab injection alone. Patients could select their preferred treatment by themselves. The study was performed in accordance with the tenets of the Declaration of Helsinki and it was approved by the Institutional Review Board of the National Taiwan University Hospital.

All patients had definite diagnosis of PDR and persistent dense VH for 3 or more weeks preventing panretinal photocoagulation (PRP). Excluded from the study were patients who had history of previous intraocular vitreoretinal surgery, such as vitrectomy or intravitreal triamcinolone injection; history of thromboembolic events, such as myocardial infarction or cerebral vascular accident; poorly controlled hypertension (systolic blood pressure above 180 mmHg or diastolic blood pressure above 110); known coagulation abnormalities or current use of anticoagulative medications other than aspirin. All cases went through ultrasonography to determine the vitreoretinal conditions. Lesion eyes with vitreoretinal point adhesion at 2 sites or less were included; those with ultrasonographic evidence of broad vitreoretinal adhesion, traction retinal detachment, or vitreoretinal point adhesion at 3 sites or more were excluded.

Demographics and clinical findings, including age, sex, metabolic control, diabetes mellitus category and duration, and previous treatments such as photocoagulation or cryotherapy, were recorded.

Baseline ophthalmologic examinations including measurement of best-corrected visual acuity (BCVA) using the Snellen acuity test, undilated and dilated slit lamp biomicroscopic examination and indirect funduscopic examination were recorded. VH severity was assessed based on the view of the optic disc and posterior pole of the retina using the indirect ophthalmoscope with a 20 diopter lens. The grading scale referred to the grading scheme of vitreous haze, ranging from 0 to 5 (0=none; 1=minimal, posterior pole clearly visible; 2=mild, posterior pole details slightly hazy; 3=moderate, posterior pole details very hazy; 4=marked, posterior pole details barely visible; 5=severe, fundal details not visible).²⁰ Post-treatment regular follow-up was done at 1 day, 1 week, and then at 3–4 weeks for 6 months. Results of ophthalmologic examinations after treatment were recorded; evidence of areas of FVP for more than 2 disc diameters was documented. Systemic or locally adverse events were monitored. Full-scatter PRP was performed when the peripheral fundus became clear. Repeated IVB were administered 4–6 weeks after the initial treatment in both groups if fundus examination showed no signs of VH decrease. If traction retinal detachment was suspected after these treatments, ocular ultrasonographic examinations were arranged during follow-up to evaluate its severity.

Pars plana vitrectomy was performed 12 to 14 weeks after initial treatment in patients with persistent or recurrent VH in both groups. Vitrectomy was also indicated if progressive FVP or macula-threatening traction, or combined detachment, developed during follow-up. Each patient was followed up at outpatient clinics for at least 6 months.

Cryotherapy and injection technique

For combined ARC and IVB procedures, patients were treated under peribulbar anesthesia with 5% xylocaine. After disinfection thrice with 5% povidone iodide and proper draping, lid speculum was put in place. The location of the first transconjunctival ARC spot was chosen where the retina was least obstructed by the blood through indirect ophthalmoscope monitoring. The foot pedal was released immediately after the appearance of white retinal patch. The remaining cryo spots were applied with the same duration as the initial spot, starting from the temporal upper quadrant. If VH was too dense for adequate visualization of any cryo patch, a fixed duration of 5 s was used for all cryo spots. After ARC, the eye was again irrigated and soaked with 5% povidone iodide. Bevacizumab (1.25 mg/0.05 mL) was drawn into a 1 mL syringe through a 27-gauge needle from a newly opened vial and injected through a 30-gauge needle into the vitreous cavity via the temporal lower pars plana. For IVB only, the patient was treated under topical anesthesia. After proper disinfection with 5% povidone iodide as above, the injection site at the temporal lower quadrant was further anesthetized with proparacaine-soaked cotton-tip applicators. Same dosage of bevacizumab as stated above was injected.

Surgical technique

For vitrectomy, standard 3-port pars plana vitrectomy was applied. Briefly, removal of opacified vitreous and anterior-posterior traction was performed first, followed by fibrovascular tissues removal with delamination as the principle technique. Hemostasis was achieved by raising the infusion bottle, applying mechanical compression with a soft-tipped cannula, endodiathermy, or a combination of the above techniques. Supplementary panretinal photocoagulation extending beyond the equator and peripheral cryotherapy were performed. Finally, cryotherapy of the sclerotomy sites (1 spot, 6 s/each, for 3 sclerotomy sites) was performed. For cases with 1 or more sites of broad vitreoretinal adhesion from FVP, fluid-air exchange with intravitreal infusion of 10% C3F8 was performed.²¹

Main outcome measures

Vitreous clear-up was defined as clearly visible main retinal vessels and the disc in the posterior pole with peripheral retina clear enough for a successful panretinal photocoagulation (PRP) in at least 3 quadrants. PRP was performed by independent ophthalmologists unaware of the procedures for initial treatment. The following results were recorded to evaluate the effects of bevacizumab: time to vitreous clear-up after initial and second treatment; frequency of second injection use; rate of persistent or recurrent VH; and indications and rate of vitrectomy during follow-up. The rate of vitreous clear-up without vitrectomy, VCUT, rate of persistent or recurrent VH, indications and rates of vitrectomy, and changes of BCVA were compared between the 2 groups. Snellen test results, converted to logarithm of minimum angle of resolution (logMAR), were used for visual acuity assessment. We assigned logMAR scores of 1.8 to count fingers, 1.9 to hand motion, and 2.0 to light perception.

Statistical analysis

All data were collected on MS-Excel 2003 spreadsheets (Microsoft Corporation) and analyzed using SPSS 13.0 for Windows (SPSS, Inc.). To examine the differences between the 2 groups, Chi-squared or Fisher's exact test (if n<5) were used. Continuous variables were presented as median or mean±standard deviation and the Mann–Whitney U test was performed. To further evaluate the effect of combined treatment and to examine other possible factors affecting VCUT in patients without vitrectomy, we performed multiple lineal regression analysis to determine the significance of the following factors: age, gender, duration of diabetes (per year), type of diabetes, prior PRP, hypertension, score of VH, and evidence of FVP, ARC with IVB, and IVB alone. We also performed multiple logistic regression analysis to examine the above possible factors in relation to final vitrectomy. For all statistical tests, P<0.05 denoted significance.

Results

The study group consisted of 35 eyes (35 patients), and the control group consisted of 32 eyes (32 patients) respectively. Demographics are summarized in Table 1. The average VH score was 4.31±0.58 for the study group and 4.25±0.75 for the controls at baseline (P=0.88). Panretinal laser therapy had been previously performed on 28 eyes (80.0%) in the study group and 26 (81.3%) in the control group (P=0.90).

Table 1.

Baseline Features of Patients With and Without Intravitreal Bevacizumab (Avastin) Injection

Characteristics	IVIA+cryotherapy (n=35)^a	IVIA only (n=32)^b	P-value
Number of eyes	35	32
Right eye	16	16	0.73
Left eye	19	16
Gender			0.27
Male	15	18
Female	20	14
Age (mean±SD)	56.9±9.2 (range 27–68)	56.1±10.4 (range 23–70)	0.61
Type of DM (n)			0.91
Type 1 (IDDM)	3	3
Type 2 (NIDDM)	32	29
Duration of DM (years) (mean±SD)	12.0±4.2	12.0±6.6	0.58
Prior PRP [n (%)]	28 (80.0)	26 (81.3)	0.90
Average score of VH	4.31±0.58	4.25±0.75	0.88
Evidence of FVP [n (%)]	15 (42.9)	14 (43.8)	0.94

Statistical analyses of age and DM duration between study and control groups were performed using Mann–Whitney U test. Other analyses were performed using Chi-squared test. With n<5, Fisher's exact test was used. P<0.05 was considered statistically significant.

IVIA+cryotherapy=patients who received intravitreal 1.25 mg bevacizumab intravitreal injection combined with anterior peripheral retinal cryotherapy.

IVIA only=patients who received intravitreal 1.25 mg bevacizumab intravitreal injection.

IVIA, intravitreal injection of Avastin; SD, standard deviation; DM, diabetes mellitus; IDDM, insulin-dependent diabetes mellitus; NIDDM, noninsulin-dependent diabetes mellitus; PRP, panretinal photocoagulation; VH, vitreous hemorrhage; FVP, fibrovascular proliferation.

In the study group, except transient conjunctival chemosis for 3 to 7 days, no cryotherapy or injection-related adverse events (either local or systemic) were observed. Four eyes had a transient increase of intraocular pressure (IOP) and it was well controlled under anti-glaucoma medications in all 4 eyes. None of the treated eyes needs long-term anti-glaucoma medications for IOP control. There was no subjective complaint of a narrowing of the horizontal visual field in all treated cases. Seven eyes (20.0%) received a second intravitreal injection of bevacizumab. Eventually, five eyes (14.3%) required vitrectomy. Over all, the VCUT from initial VH was 7.8±9.0 weeks (median: 4.1 weeks). Excluding vitrectomy cases, the VCUT was 6.9±8.7 weeks (median: 4.1 weeks). Among those 5 cases, surgical indications were for dense VH, refractory to cryotherapy, and repeated injection of bevacizumab in 3 eyes (60%) and traction retinal detachment in 2 eyes (40%).

In the control group, 8 out of 32 eyes (25.0%) underwent second injection. Four eyes (12.5%) eventually received pars plana vitrectomy. Over all, the VCUT was 13.5±8.9 weeks (median: 11.3 weeks) in cases with eventual vitreous blood reabsorption. Excluding vitrectomy cases, the VCUT was 13.0±9.3 weeks (median: 9.6 weeks) in the remaining 28 eyes (87.5%). Among these 4 vitrectomized eyes, surgical indications pointed to dense VH in 2 eyes (50%), severe FVP in 1 eye, and traction retinal detachment in 1 eye after vitreous clear-up. The clinical outcomes of the study and the control groups were summarized in Table 2. Because of the small case number operated on in the study group, the degree of fibrosis, severity of intraoperative bleeding or area of detachment between 2 groups could not be adequately compared.

Table 2.

Clinical Outcomes for Patients With and Without Intravitreal Bevacizumab (Avastin) Injection

Clinical outcome	IVIA+cryotherapy^a (n=35)	IVIA only^b (n=32)	P-value
Vitreous clear-up time^c (weeks) [mean±SD] (median)	7.8±9.0 (4.1)	13.5±8.9 (11.3)	0.003
Vitreous clear-up time without vitrectomy (weeks) [mean±SD] (median)	6.9±8.7 (4.1)	13.0±9.3 (9.6)	0.003
Repeated injection of bevacizumab [n (%)]	7 (20.0)	8 (25.0)	0.62
Eyes underwent vitrectomy during follow-up [n (%)]	5 (14.3)	4 (12.5)	0.83
Indications for vitrectomy [n (%)]			1.0
Dense VH	3 (8.6)	2 (6.3)
Combined VH and FVP	0 (0)	1 (3.1)
Combined VH and TRD	2 (5.7)	1 (3.1)
Eyes with nonclearing VH within 12 months [n (%)]	0 (0)	0 (0)	—

P<0.05 was considered statistically significant; statistical comparisons between groups were performed using Chi-square/Fisher's exact test or Kruskal–Wallis test (for vitreous clear-time and duration of VH).

IVIA+cryotherapy=patients who received intravitreal 1.25 mg bevacizumab intravitreal injection combined with anterior peripheral retinal cryotherapy.

IVIA only=patients who received intravitreal 1.25 mg bevacizumab intravitreal injection.

All cases with or without vitrectomy.

TRD, traction retinal detachment.

The VCUT between the study and the control groups were significantly different, whether vitrectomy cases were excluded or not. (P=0.03 with vitrectomy cases included and 0.003 with vitrectomy cases excluded.) However, there was no difference in the vitrectomy rates between these 2 groups (14.3% in study group, 12.5% in control group, P=0.83).

At baseline, mean BCVA was 1.73±0.35 in logMAR in the study group, and 1.34±0.60 in logMAR in the control group (P=0.02) (Table 3). Compared to the study group, the control group has a better baseline visual acuity before treatment. BCVA 6 months following the first treatment improved to 0.81±0.48 in logMAR in the study group and to 0.98±0.50 in logMAR in the control group (P=0.91). The difference in initial and post-operative 6 month BCVA (logMAR) between the study and control groups was statistically significant (P=0.01).

Table 3.

Best-Corrected Visual Acuity for the Study and Control Groups

	IVIA+cryotherapy^a (n=35)	IVIA only^b (n=32)	P-value
Baseline BCVA (logMAR)	1.73±0.35	1.34±0.60	0.02
Postop 6 months BCVA (logMAR)	0.89±0.51	0.89±0.51	0.91
Difference between baseline and postop 6 months BCVA (logMAR)	−0.84±0.53	−0.45±0.58	0.01

Statistical analyses between study and control groups were performed using Mann–Whitney test. P<0.05 was considered statistically significant.

IVIA+cryotherapy=patients who received intravitreal 1.25 mg bevacizumab intravitreal injection combined with anterior peripheral retinal cryotherapy.

IVIA only=patients who received intravitreal 1.25 mg bevacizumab intravitreal injection. We assigned logMAR scores of 1.8 to count fingers, 1.9 to hand motion, and 2.0 to light perception.

BCVA, best-corrected visual acuity; LogMAR, logarithm of minimum angle of resolution.

The results of multiple lineal regression analysis revealed that combined treatment with cryotherapy and bevacizumab would shorten the VCUT compared with the bevacizumab monotherapy (95%CI −10.74 to −0.86, P=0.02).

Discussion

Panretinal photocoagulation has been the main treatment method for PDR over the past several decades.^2–4 However, in patients with media opacity such as VH, it is not always possible to administer full-scattered panretinal photocoagulation to inhibit continuous development of neovascularization. Our previous study demonstrates IVB may shorten the duration of VH¹⁹; in this study, combined therapy of IVB and transconjunctival panretinal cryotherapy were used to compare its effects on vitreous clear-up with that using IVB only. The mean duration of vitreous opacity was shorter in the study group than in the control bevacizumab monotherapy group. Multiple regression analysis confirms the association of bevacizumab injection and peripheral retinal cryotherapy with reduced VCUT. These results suggest that the combined use of intravitreal injection of bevacizumab and panretinal cryotherapy may be an effective combination therapy for PDR with VH.

In diabetic retinopathy, prolonged VH may be caused by too much blood exceeding the rate of normal reabsorption process; alternatively, it may be caused by recurrent bleeding from the initial bleeding site and fresh bleeding from different sites of neovascularization during the process of blood reabsorption. Further, FVP may develop under the coverage of the blood, caused by stimulation by various growth factors contained within the blood or released from the ischemic retina, thus forming a vicious circle. IVB may induce temporary regression of retinal neovascularization and cessation of hemorrhage from all potential bleeding sources, breaking the neovascularization stimulation process.^22–25 However, it has no effect on enhancing the reabsorption of dense preexisting blood; thus, this treatment may still leave some patients with prolonged reduced vision.

Panretinal cryotherapy has long been used as a substitute for PRP to treat PDR when dense VH is present. It has also been used as supplementary treatment when full scattered PRP fails to inhibit growth of retinal neovascularization. In eyes undergoing diabetic vitrectomy, peripheral retinal cryotherapy has been shown to reduce the rate of postoperative VH.¹⁰ The mechanism of cryotherapy to prevent the occurrence or shorten the duration of VH may be 2-fold: first, similar to PRP, the treatment may decrease the production of vascular endothelial growth factor (VEGF) by inducing oxygen redistribution or stimulation of anti-VEGF release^6–8; second, the treatment causes sufficient inflammation, facilitating the recruitment of macrophage, which in turn help carrying out red blood cells or blood components.¹⁷ The increased inflammation has the undesirable effect of increasing the vascular permeability and stimulating the production of growth factors, further stimulating neovascularization. This may be part of the reason why cryotherpy alone is not an efficient way of treating nonclearing VH. With the combined use of bevacizumab, the most important neovascularization promoting factor–VEGF can be effectively counterbalanced, the side effect of cryotherpy may be reduced, and its blood reabsorption effect may thus be enhanced.

Our treatment technique involved performing the cryotherapy through conjunctiva under peribulbar anesthesia; 12 spots were applied to 360 degree of the peripheral retina. Temporal side was treated first because of its easy access; nasal side became more easily accessible as IOP gradually decreased. The treatment duration was set by direct monitoring of the first spot or a fixed 5 s. The latter was decided from our previous experiences when cryotherpy was performed with clear media. With our technique, the patients generally felt no pain, and there might be mild intraocular inflammation and a few days of conjunctival chemosis. Our study suggests that while combined treatment may shorten blood reabsorption time, the rate of eventual vitrectomy may not be reduced compared with bevacizumab injection alone. Nonetheless, the faster vitreous clear-up remains a major advantage for the patient especially if the fellow eye has reduced vision. Our retrospective study did not have perfect match between the study and the control group. The preoperative visual acuity in the study group was worse than the control group, and the study group showed more advanced VH. The postoperative vision of the study group, however, was as good as that of the control group. Thus, these data seem to strengthen rather than diminish the power of our study.

In our study, increased traction from preexisting fibrovascular tissue is a major concern. In theory, the use of anti-VEGF might tip the balance among different growth factors toward fibrosis-inducing factors and raise the risk of enhanced vitreoretinal traction or traction retinal detachment; panretinal cryotherapy may also induce vitreous contraction. Therefore, it is crucial to perform ultrasonography to exclude cases with significant vitreoretinal adhesion. In this study, 2 cases developed traction detachment after combined treatment. Although it could not be certain that this complication was related to the treatment, caution should be exercised to monitor the fundus changes by careful examination and ultrasonographic follow-up. Another concern of the cryotherapy is the possibility of destruction of the peripheral retina and the loss of peripheral visual field. Our technique used transconjunctival approach and thus the postequatorial area was not treated. This approach may minimize the visual field loss. Cryotherapy may also affect choroidal circulation in the treated area. The treatment should not be excess in individual spot duration and spot number. We believe our approach with direct monitoring retinal changes during cryotherapy and the three spots per quadrant dosage has reduced the adverse effect as much as possible.

Our study had limitation and bias, including the small case number, retrospective nature, nonrandomized patients section and varied follow-up period. The 2 groups in our study were not matched in the category of VH severity grading, with the combined treatment group presenting with more severe hemorrhage and worse initial visual acuity. However, as mentioned before, these data did not weaken our conclusion, and may even strengthen it. Our study results suggest that in PDR with VH, combined intravitreal injections of bevacizumab and panretinal cryotherapy may be associated with rapid vitreous clearance, and improvement in visual acuity compared with bevacizumab monotherapy. Further prospective randomized study may be required to confirm our results, and to determine the optimal cryo spots and injection interval.

Footnotes

Acknowledgment

This research was supported by the Ophthalmology Departmental Core Grand of the National Taiwan University Hospital.

Author Disclosure Statement

The authors have no commercial or financial interests associated with this article.

References

Michels

R.G.

Proliferative diabetic retinopathy: pathophysiology of extraretinal complications and principles of vitreous surgery. Retina. 1:1–17, 1981.

Preliminary report on effects of photocoagulation therapy. The Diabetic Retinopathy Study Research Group. Am. J. Ophthalmol. 81:383–396, 1976.

Photocoagulation treatment of proliferative diabetic retinopathy. Clinical application of Diabetic Retinopathy Study (DRS) findings, DRS Report Number 8. The Diabetic Retinopathy Study Research Group. Ophthalmology. 88:583–600, 1981.

Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology. 98:766–785, 1991.

Shimura

, Yasuda

, Nakazawa

, Kano

, Ohta

, and Tamai

Quantifying alterations of macular thickness before and after panretinal photocoagulation in patients with severe diabetic retinopathy and good vision. Ophthalmology. 110:2386–2394, 2003.

Schimek

R.A.

, and Spencer

Cryopexy treatment of proliferative diabetic retinopathy. Retinal cryoablation in patients with severe vitreous hemorrhage. Arch Ophthalmol., 97:1276–1280, 1979.

Mosier

M.A.

, Del Piero

, and Gheewala

S.M.

Anterior retinal cryotherapy in diabetic vitreous hemorrhage. Am. J. Ophthalmol., 100:440–444, 1985.

Benedett

, Olk

R.J.

, Arribas

N.P.

, Okun

, Johnston

G.P.

, Boniuk

, Escoffery

R.F.

, Grand

M.G.

, and Schoch

L.H.

Transconjunctival anterior retinal cryotherapy for proliferative diabetic retinopathy. Ophthalmology. 94:612–619, 1987.

Neely

K.A.

, Scroggs

M.W.

, and McCuen

B.W.

2nd.

Peripheral retinal cryotherapy for postvitrectomy diabetic vitreous hemorrhage in phakic eyes. Am. J. Ophthalmol., 126:82–90, 1998.

10.

Yeh

P.T.

, Yang

C.M.

, Yang

C.H.

, and Huang

J.S.

Cryotherapy of the anterior retina and sclerotomy sites in diabetic vitrectomy to prevent recurrent vitreous hemorrhage: an ultrasound biomicroscopy study. Ophthalmology. 112:2095–2102, 2005.

11.

Chen

, and Park

C.H.

Use of intravitreal bevacizumab as a preoperative adjunct for tractional retinal detachment repair in severe proliferative diabetic retinopathy. Retina. 26:699–700, 2006.

12.

Mason

J.O.

3rd , Nixon

P.A.

, and White

M.F.

Intravitreal injection of bevacizumab (Avastin) as adjunctive treatment of proliferative diabetic retinopathy. Am. J. Ophthalmol., 142:685–688, 2006.

13.

Avery

R.L.

, Pearlman

, Pieramici

D.J.

, Rabena

M.D.

, Castellarin

A.A.

, Nasir

M.A.

, Giust

M.J.

, Wendel

, and Patel

Intravitreal bevacizumab (Avastin) in the treatment of proliferative diabetic retinopathy. Ophthalmology. 113:1695. e1691–1615, 2006.

14.

Spaide

R.F.

, and Fisher

Y.L.

Intravitreal bevacizumab (Avastin) treatment of proliferative diabetic retinopathy complicated by vitreous hemorrhage. Retina. 26:275–278, 2006.

15.

Huang

Y.H.

, Yeh

P.T.

, Chen

M.S.

, Yang

C.H.

, and Yang

C.M.

Intravitreal bevacizumab and panretinal photocoagulation for proliferative diabetic retinopathy associated with vitreous hemorrhage. Retina. 29:1134–1140, 2009.

16.

Curtin

V.T.

, Fujino

, and Norton

E.W.

Comparative histopathology of cryosurgery and photocoagulation. Observations on the advantages of cryosurgery in retinal detachment operations. Arch. Ophthalmol., 75:674–682, 1966.

17.

Williams

D.F.

, Burke

J.M.

, and Williams

G.A.

Clearance of experimental vitreous hemorrhage after panretinal cryotherapy is related to macrophage influx. Arch. Ophthalmol., 108:595–597, 1990.

18.

Cekic

, and Batman

Peripheral retinal cryotherapy for postvitrectomy diabetic vitreous hemorrhage in phakic patients. Am. J. Ophthalmol., 127:740–741, 1999.

19.

Yeh

P.T.

, Yang

C.H.

, and Yang

C.M.

Intravitreal bevacizumab injection for recurrent vitreous haemorrhage after diabetic vitrectomy. Acta Ophthalmol., 89:634–640, 2011.

20.

Nussenblatt

R.B.

, Palestine

A.G.

, Chan

C.C.

, and Roberge

Standardization of vitreal inflammatory activity in intermediate and posterior uveitis. Ophthalmology. 92:467–471, 1985.

21.

Yang

C.M.

, Yeh

P.T.

, Yang

C.H.

, and Chen

M.S.

Bevacizumab pretreatment and long-acting gas infusion on vitreous clear-up after diabetic vitrectomy. Am. J. Ophthalmol., 146:211–217, 2008.

22.

Jorge

, Costa

R.A.

, Calucci

, Cintra

L.P.

, and Scott

I.U.

Intravitreal bevacizumab (Avastin) for persistent new vessels in diabetic retinopathy (IBEPE study). Retina. 26:1006–1013, 2006.

23.

Shih

C.W.

, Yang

C.M.

, Chen

M.S.

, and Wang

T.J.

Intravitreal injection of bevacizumab and gas for diabetic premacular hemorrhage with active fibrovascular proliferation. Graefes Arch. Clin. Exp. Ophthalmol., 246:1547–1551, 2008.

24.

Avery

R.L.

Regression of retinal and iris neovascularization after intravitreal bevacizumab (Avastin) treatment. Retina. 26:352–354, 2006.

25.

Isaacs

T.W.

, and Barry

Rapid resolution of severe disc new vessels in proliferative diabetic retinopathy following a single intravitreal injection of bevacizumab (Avastin). Clin. Exp. Ophthalmol., 34:802–803, 2006.