Acta Ophthalmologica 2010

Bevacizumab as adjuvant for

neovascular glaucoma
Julia Beutel,1 Swaantje Peters,1 Matthias Lüke,1 Sabin Aisenbrey,2
Peter Szurman,2 Martin S. Spitzer,2 Efdal Yoeruek2, the
Bevacizumab Study Group and Salvatore Grisanti1
1
Department of Ophthalmology, University of Lübeck, Lübeck, Germany
2
University Eye Hospital, Centre for Ophthalmology, Eberhard-Karls University of
Tuebingen, Tuebingen, Germany

ABSTRACT. Introduction
Purpose: We aimed to evaluate the longterm effects of intraocular bev- Proliferative ischaemic retinopathies
acizumab (Avastin) injections as adjuvant treatment in patients with neovas- with increased levels of vascular endo-
cular glaucoma. thelial growth factor (VEGF) lead to
Methods: Twenty eyes of 18 consecutive patients with secondary neovascular abnormal vessels on the iris called
glaucoma caused by proliferative diabetic retinopathy (n = 7), ischaemic cen- rubeosis. In progressive stages, fibro-
tral retinal vein occlusion (n = 7), ischaemic ophthalmopathy (n = 2) and ret- vascular membranes may develop,
inal ischaemia resulting from persistent detachment (n = 2) were treated with occlude the anterior chamber angle
intraocular bevacizumab injections (1.25 mg ⁄ 0.05 ml) in addition to other and inhibit aqueous outflow, resulting
treatments. The main outcome measure was the change in degree of iris rubeo- in neovascular glaucoma. Established
treatment modalities include panretinal
sis. Secondary outcomes included intraocular pressure (IOP), best corrected
photocoagulation to ablate the source
visual acuity (BCVA) and numbers of additional interventions or antiglaucoma
of elevated VEGF, and cyclodestruc-
medications administered after injection. tive or drainage procedures to counter-
Results: Mean (± standard deviation) follow-up was 67.7 ± 13.8 weeks act the angular obstruction and reduce
(range 50–93 weeks). At the last follow-up, complete regression of rubeosis intraocular pressure (IOP). A recently
was detectable in five (20%) eyes, incomplete regression in seven (35%), sta- proposed therapeutic option is the
bilization in six (30%), and an increase in two (10%) eyes. Mean IOP was additional inhibition of the angiogenic
26.0 ± 8.9 mmHg at baseline and significantly decreased to 14.75 ± 5.3 cascade by intraocularly injected bev-
mmHg at the last follow-up visit (p = 0.000005). Mean baseline BCVA acizumab (Avastin; Genentech, Inc.,
(logMAR [logarithm of the minimum angle of resolution] 1.43 ± 0.89) was South San Francisco, CA, USA). Pre-
stabilized during the follow-up period (logMAR 1.5 ± 0.98). Patients received vious short-term case reports have
an average of 2.75 injections. Additional treatments were laser photocoagu- described the acute effects based on
lation in 13 (65%) eyes, cyclodestructive procedure in 14 (70%), cryopexy in reduction of vascular leakage and reg-
six (30%), drainage procedures in two (10%), and vitrectomy in five (25%) ression of anterior segment neovascu-
eyes. larization (Avery 2006; Davidorf et al.
Conclusions: Bevacizumab may be beneficial as adjuvant treatment in neovas- 2006; Grisanti et al. 2006; Iliev et al.
2006; Kahook et al. 2006; Mason
cular glaucoma because of its anti-angiogenic properties and its ability to pre-
et al. 2006; Silva Paula et al. 2006).
vent establishment or progression of angular obstruction. The causative disease
The aim of this study was to report the
inducing the angiogenic process requires treatment in all cases. Antiglaucoma
longterm outcome after adjuvant intra-
treatment is needed in cases of persistent elevated IOP. ocular bevacizumab application in
patients with neovascular glaucoma
Key words: bevacizumab – iris rubeosis – ischaemic retinopathy – neovascular glaucoma – and to analyse the role of bevacizumab
VEGF
within a holistic treatment regime.

Acta Ophthalmol. 2010: 88: 103–109

ª 2008 The Authors Materials and Methods
Journal compilation ª 2008 Acta Ophthalmol
We retrospectively reviewed 52 eyes of
doi: 10.1111/j.1755-3768.2008.01355.x 50 patients with neovascular glaucoma

103
Acta Ophthalmologica 2010
who received treatment at the Univer-
sity Eye Hospital Tübingen between
October 2005 and September 2006.
Inclusion criteria were neovascular
glaucoma, rubeosis, follow-up data
for 12 months, written informed con-
sent and visual acuity (VA) of light
perception or better. In cases of
uncontrolled hypertension, or in cases
with a history of thromboembolic
events including myocardial infarc-
tion, and cerebral insult and renal dis-
ease, no off-label use of intravitreal
Avastin was offered. Thirty-two eyes
of 32 patients were excluded (Fig. 1).
Inclusion and exclusion criteria are
given in Table 1. Informed consent
was obtained after detailed informa-
tion about the off-label nature of the
intraocular bevacizumab injections,
and the potential risks of endoph-
thalmitis and retinal detachment, as
well as potential repetition of the
treatment.
Fig. 1. Flow chart showing patient character-
istics.
Table 1. Inclusion and exclusion criteria.
Inclusion criteria
Neovascular glaucoma
Rubeosis
Available 12-month follow-up
Written informed consent
Visual acuity of light perception or better
Exclusion criteria
Uncontrolled hypertension
Renal disease
History of thromboembolic events
(including myocardial infarction, cerebral
insult)
104
The majority of eyes (16 of 20) had
received at least one treatment for
neovascular glaucoma at baseline
(Table 2). Histories of previous laser
photocoagulation, cyclodestructive
procedures and cryopexy were present
for 14 (70%), nine (45%) and six
(30%) eyes, respectively. Pars plana
vitrectomy had been previously per-
formed in eight (40%) eyes at base-
line. Three (15%) patients had
received triamcinolone acetonide injec-
tion previously. At baseline, all
patients were using at least one topical
antiglaucoma medication (2.8 ±
1.01). Additionally, 10 (50%) patients
were using systemic acetazolamide.
Baseline characteristics of the patients
are summarized in Table 2.
Complete ophthalmic examinations
were performed at baseline and at fol-
low-up during the first 2 weeks
(9.7 ± 4.7 days, range 2–18 days),
after 1–2 months (43.5 ± 20.7 days,
range 29–119 days), after 6 months
(24.9 ± 1.9 weeks, 21–19 weeks) and
after 12 months (54.6 ± 5.0 weeks,
range 48–70 weeks). These included
best corrected visual acuity (BCVA),
slit-lamp examination, and IOP mea-
surements with the Goldmann appla-
nation tonometer and gonioscopy. Iris
fluorescein angiography (FA) was
obtained in most cases (n = 15).
Grading of iris FA was performed
according to the classification of Eh-
renberg et al. (1984) (Table 3). In the
remaining five cases, rubeosis visible
at slit-lamp examination was graded
according to the same guidelines. Best
corrected VA was measured with ET-
DRS charts at 4 m.
Re-injections were performed at the
discretion of the investigators. In gen-
eral, bevacizumab injections were
repeated if a persistent rubeosis
detected by iris FA or slit-lamp
examination and elevated IOP were
evident. The injection was combined
with laser photocoagulation depend-
ing on the degree of ischaemia accord-
ing to FA and the underlying cause of
ischaemia (Diabetic Retinopathy
Study 1978; Eggleston et al. 1980;
Central Vein Occlusion Study Group
1995) or with peripheral retinal
cryopexy. In cases of uncontrolled
IOP, a combination with a cyclode-
structive procedure (cyclophotocoagu-
lation, cyclocryocoagulation) was
chosen to achieve rapid and prolonged
IOP control.
Injections were performed in the
operating theatre under standardized
sterile conditions. For disinfection,
povidone iodine was applied directly
to the eyelid margins, eyelashes and
conjunctival ocular surface. Topical
anaesthesia was applied and a lid
speculum inserted, after which
patients received an intravitreal or
intracameral injection of a 0.5 ml vol-
ume containing 1.25 mg bevacizumab
(Avastin; Genentech Inc.). The main
outcome measure was the change in
degree of rubeosis at 12 months fol-
low-up. Secondary outcomes included
IOP, BCVA, numbers of additional
interventions or antiglaucoma medica-
tions after injection.
The study protocol adhered to the
guidelines of the Declaration of Hel-
sinki as revised in Tokyo and Venice.
For statistical analysis, origin Version
6.0 (Microcal Software, Inc., North-
ampton, MA, USA) was used. Signifi-
cance was estimated by Student’s
t-test and p £ 0.05 was considered sta-
tistically significant.
Results
In total, we reviewed 20 eyes in 18
consecutive patients receiving bev-
acizumab for neovascular glaucoma
caused by proliferative diabetic reti-
nopathy (n = 7), ischaemic retinal
vein occlusion (n = 7), ischaemic oph-
thalmopathy (n = 2), and retinal
ischaemia caused by detachment
(n = 2). Of the last two patients, one
presented after cerclage surgery with a
persisting retinal detachment localized
anterior to the cerclage (patient 13).
The other had a prolonged tractional
retinal detachment (patient 18). Fif-
teen of the 18 patients (20 eyes) were
male and three female. The duration
of follow-up was 67.7 ± 13.8 weeks
(range 50–93 weeks). Mean age was
68.9 years (range 55–90 years). The 20
eyes received an average of 2.75 injec-
tions each (11 intracameral, 44 intra-
vitreal). Repeated injections were
performed in eight (40%) eyes. Patient
characteristics are shown in Table 2.
At baseline, iris FA (performed in
15 eyes) revealed rubeosis grades 1
and 2 in one (7%) eye, respectively.
Five (33%) eyes had rubeosis grade 3
and eight (53%) eyes presented with
rubeosis grade 4. At the last follow-
up, no rubeosis was detectable in five
(20%) eyes. Regression was seen in
 Acta Ophthalmologica 2010

Table 2. Baseline characteristics.

Avastin Local Systemic Grading

Patient Age, Previous treatments, Systemic glaucoma acetazolamide, of iris IOP, BCVA,
no. years Eye Diagnosis therapies n disorder medications, n mg ⁄ day rubeosis* mmHg logMAR

1 69 R DR LP 2 AH, DM 4 750 3 34 1.1

L LP, CDP, CP 2 AH, DM 4 750 4 34 0.6
2 69 R DR LP 1 AH, DM 4 500 3 14 0.2
L LP, CDP 1 AH, DM 1 250 4 18 0.4
3 58 L CRVO VA 1 AH, DM 3 750 4 24 1.8
4 75 R CRVO LP, CDP, ppV, CP 1 AH 3 375 4 32 1.2
5 56 R DR LP, CDP, 4 AH, DM 3 0 4 23 1.8
VA, ppV, CP
6 73 R CRVO ) 1 AH, DM 3 750 3 33 2.3
7 76 R IO LP, CDP, CP 1 AH 4 750 4 27 1.8
8 65 R DR ) 1 AH 4 0 1 40 3.0
9 69 L BRVO LP, ppV 5 AH 1 0 4 16 0.6
10 75 L DR LP, CDP, CP 1 AH, DM 3 0 3 23 0.5
11 69 R IO LP, ppV, CP 1 AH 3 0 ) 34 0.7
12 57 R DR LP, CDP, ppV 4 AH, DM 3 750 ) 43 1.3
13 79 R PD ) 1 AH 2 0 3 30 3.0
14 72 L CRVO LP, CDP, VA, ppV 21 AH 2 0 4 16 1.0
15 55 L DR LP, ppV 3 DM 2 0 ) 10 0.7
16 78 R CRVO CDP 1 AH 1 500 ) 29 3.0
17 90 L CRVO ) 1 AH 3 0 ) 32 1.8
18 54 L PD LP, ppV 2 AH 3 0 2 26 1.8

* According to Ehrenberg et al. (1984).

IOP = intraocular pressure; BCVA = best corrected visual acuity; R = right eye; L = left eye; DR = diabetic retinopathy; CRVO = central
retinal vein occlusion; IO = ischaemic ophthalmopathy; BRVO = branch retinal vein occlusion; PD = persistent retinal detachment; LP = laser
photocoagulation; CDP = cyclodestructive procedure; CP = cryopexy; ppV = pars plana vitrectomy; VA = volon-A injection; AH = arterial
hypertension; DM = diabetes mellitus.

Table 3. Grading of iris fluorescein angio-
graphs according to the classification of
Ehrenberg et al. (1984).
Grade Angiographic aspect
0 No fluorescein leakage
1 Mild leakage in one or two
quadrants of the pupillary sphincter
2 Mild leakage in three or four
quadrants of the pupillary sphincter
3 Pupillary sphincter leakage in
three or four quadrants combined
with leakage in one or two quadrants
of the iris stroma
4 Pupillary sphincter leakage in three
or four quadrants combined
with iris stromal leakage
in at least three quadrants
seven (35%) eyes. An unchanged
degree of rubeosis was seen in six
(30%) eyes and an increase of rubeosis
in two (10%). Evaluation by iris FA
revealed complete regression of rubeo-
sis in two (13%) eyes after bevaci-
zumab injection. Grade 1 rubeosis was
detected in two (13%) eyes. Five
(33%) patients presented with rubeosis
grade 2. Grades 3 and 4 were evident
in three (20%) patients, respectively.
Involvement of the anterior chamber
angle was evaluated by gonioscopy in
10 patients (Table 4).
Mean IOP at baseline was 26.0 ±
8.9 mmHg (range 10–43 mmHg), 95%
confidence interval [CI] 22.8–31.0
mmHg) and decreased to 19.1 ±
9.3 mmHg (range 8–36 mmHg) during
the first 2 weeks, to 19.0 ± 9.3
mmHg (range 8–36 mmHg) after
2 months, and to 16.47 ± 5.1 mmHg
(range 10–31 mmHg) at 6 months. At
12 months after the first injection,
IOP was 14.75 ± 5.3 mmHg (range
6–28 mmHg, 95% CI 12.3–17.2
mmHg). Reduction of IOP was signifi-
cant during the first 2 weeks (p =
0.006), at month 2 (p = 0.0095),
month 6 (p = 0.00014) and month 12
after the first injection (p =
0.0000005) (Fig. 2, Table 5).
At baseline mean VA was 10 ⁄ 250
(logMAR 1.43 ± 0.89). Two weeks
and 2 months after the initial bev-
acizumab injection, mean VA was
10 ⁄ 330 (logMAR 1.47 ± 0.92) and
10 ⁄ 250 (1.41 ± 1.01), respectively.
After 6 and 12 months, BCVA was
10 ⁄ 200 (logMAR 1.28 ± 0.9) and
10 ⁄ 330 (logMAR 1.5 ± 0.98), respec-
tively. Visual acuity remained stable
with a slight reduction over the entire
follow-up period (Fig. 3). No signifi-
cant changes in VA were evident

Table 4. Gonioscopic findings of the anterior chamber angle (n = 10).

Incomplete synechia Complete

Parameter No verifiable rubeosis (n = 2) Rubeosis (n = 3) (n = 4) synechia (n = 1)

Mean IOP reduction, mmHg ) 1.5 ) 5.0 ) 13.0 ) 5.0

Additional treatments LP (n = 2), CDP (n = 1), CDP (n = 1), none LP (n = 4), CDP LP and CDP
ppV (n = 1), CP (n = 1) (n = 2) (n = 3), ppV (n = 3)
Change in degree of rubeosis Regression (n = 2) Increase (n = 1), Unchanged (n = 2), Regression
regression (n = 2) regression (n = 2)

IOP = intraocular pressure; LP = laser photocoagulation; CDP = cyclodestructive procedure; ppV = pars plana vitrectomy; CP = cryopexy.

105
Acta Ophthalmologica 2010

during the first 2 weeks (p = 0.889),

1–2 months (p = 0.963), 6 months
(p = 0.628) and 12 months (p =
0.737) (Fig. 3, Table 5).
In 13 eyes, combination therapy
was conducted and a second proce-
dure was performed within 1 week
after bevacizumab injection. A repre-
sentative case is shown in Fig. 4. Nine
of these eyes (69%) showed a reduc-
tion in IOP to < 21 mmHg during
the first 2 weeks of follow-up. In
seven eyes, further therapies were
applied during the subsequent course
of disease. In four of these eyes
(57%), IOP decreased to < 21 mmHg
Fig. 2. Boxplots showing intraocular pressure (IOP, mmHg) at baseline, week 2, month 2, in the first 2 weeks after injection.
month 6 and month 12 (p < 0.05, two-sided t-test).
Overall, additional laser photocoagu-
lation was performed in 13 (65%)
eyes and cryopexy in six (30%) eyes.
Fourteen (70%) eyes underwent an
adjunctive cyclodestructive procedure
which had to be repeated in five
(20%) eyes at least once. Drainage
procedures were performed in two
(10%) eyes. Five (25%) eyes received
repeated vitrectomies for recurrent
or persisting vitreous haemorrhage
and ⁄ or proliferative vitreoretinopathy.
Three patients did not receive further
antiglaucoma therapy. At the last
follow-up, local antiglaucoma ther-
apy was reduced or discontinued in
10 (50%) patients. A mean of 2.4 ±
Fig. 3. Boxplots showing best corrected visual acuity (BCVA, logMAR) at baseline, week 2,
month 2, month 6 and month 12 (p > 0.05, two-sided t-test). 1.4 different topical antiglaucoma

Table 5. Clinical course.

IOP at Grading IOP BCVA at BCVA at

Patient Avastin Intra- Intra- 6 months, of iris at 12 months. 6 months, 12 months,
no. injections, n cameral vitreal Further therapies mmHg rubeosis* mmHg logMAR logMAR

1 2 1 1 LP, CDP, DP, CP 13 2 13 0.8 0.7

2 1 1 LP, CDP 23 4 13 0.6 0.6
2 1 0 1 ) 16 4 20 0.0 0.2
1 1 0 LP, CDP 11 3 20 0.4 0.3
3 1 1 0 LP, CDP, ppV ) 1 12 ) 1.8
4 1 0 1 LP, CDP, ppV 17 1 10 1.4 2.6
5 4 1 3 LP, ppV 16 4 13 1.3 1.4
6 1 0 1 LP, CDP 16 3 28 2.3 2.3
7 1 0 1 CDP 19 3 12 3.0 3.0
8 1 0 1 CDP, CP 19 2 8 3.0 3.0
9 5 0 5 ) 19 2 20 0.5 1.0
10 1 0 1 LP, CDP 17 2 18 0.4 0.3
11 1 0 1 CDP 31 ) 9 0.5 1.0
12 4 1 3 LP, CDP, DP, CP 11 ) 16 1.2 1.3
13 1 0 1 LP, CDP, CP ) 0 21 ) 3.0
14 21 0 21 LP, CP 12 0 15 0.8 0.8
15 3 2 1 LP, ppV 12 ) 16 0.9 1.3
16 1 0 1 CDP, CP 18 ) 6 3.0 3.0
17 1 1 0 LP, CDP, ppV 10 ) 10 1.2 1.2
18 2 2 0 ) ) 2 15 1.8 1.8

* According to the classification of Ehrenberg et al. (1984).

IOP = intraocular pressure; BCVA = best corrected visual acuity; LP = laser photocoagulation; CDP = cyclodestructive procedure;
DP = drainage procedure; CP = cryopexy; ppV = pars plana vitrectomy.

106
 Acta Ophthalmologica 2010

Fig. 4. A 69-year-old man was referred with a

history of rubeosis secondary to diabetic reti-
nopathy in the left eye. (A) Ten and (B)
120 seconds before bevacizumab injection.
Iris angiography of the same eye at 4 days
after bevacizumab injection and 1 day after
additional panretinal photocoagulation. (C,
D) At 1 month after bevacizumab injection
and panretinal photocoagulation, cyclophoto-
coagulation was performed to reduce intra-
ocular pressure (IOP). (E, F) At 12 months,
the patient presented with stable IOP (G, H).

(A) (B) such as endophthalmitis, inflamma-

tion, increased IOP, retinal tear and
detachment, or systemic complications
such as thromboembolic events were
observed during a follow-up of
‡ 1 year.

Discussion
The mechanism of anti-VEGF therapy
in neovascular glaucoma is based on
the inhibition of neovascular activity
and vascular permeability. Bevacizu-
mab, a recombinant humanized anti-
VEGF antibody, binds to all isoforms
(C) (D)
of VEGF-A (Ferrara et al. 2004) and
seems to be relatively safe after intra-
ocular injection (Kernt et al. 2007;
Shima et al. 2007). In a recent study,
we demonstrated the dramatic short-
term effects of bevacizumab on the
vessels, evident in the fast reduction
of leakage seen in iris FA and the
regression of iris and angle neovascu-
larization. Similarly, other studies
have reported a regression of iris and
angle neovascularization secondary to
diabetic retinopathy or vein occlusion
after treatment with bevacizumab
(E) (F) (Davidorf et al. 2006; Grisanti et al.
2006; Iliev et al. 2006; Kahook et al.
2006; Mason et al. 2006; Silva Paula
et al. 2006).
These studies were all limited by
short follow-up periods, ranging
between 1 week and 4 months, and
small samples of a maximum of six
patients (Davidorf et al. 2006; Grisanti
et al. 2006; Iliev et al. 2006; Kahook
et al. 2006; Mason et al. 2006; Silva
Paula et al. 2006). The conclusions
they drew concerning both the safety,
dosage and frequency of anti-VEGF
treatment in neovascular glaucoma,
(G) (H) and the selection and timing of
additional treatment modalities, were
medications were applied. Aceta- No case of severe secondary angle- limited. In our study, we repeatedly
zolamide was not necessary in any closure glaucoma occurred during the showed that rubeosis of the iris second-
patient. follow-up. No other complications ary to a variety of ischaemic diseases

107
Acta Ophthalmologica 2010
responds to anti-VEGF therapy. Addi-
tionally, no other local complications
or systemic side-effects were observed
during a follow-up of ‡ 1 year.
Planning the optimal therapeutic
approach requires some familiarity
with the pharmacokinetics of bev-
acizumab. The first signs of an
anti-angiogenic effect have been
documented about 48 hours after
intracameral or intravitreal injection
(Grisanti et al. 2006; Iliev et al. 2006).
These effects persisted independently
of the underlying disease for
‡ 4 weeks after intracameral injection
(Grisanti et al. 2006) and 6–8 weeks
after intravitreal injection (Iliev et al.
2006; Kahook et al. 2006; Silva Paula
et al. 2006). In our first cases, the
injections were performed into the
anterior chamber to obtain an imme-
diate effect at the targeted site
(Grisanti et al. 2006). We recognized,
however, that an intravitreal injection
may as well treat the anterior and pos-
terior pathology (Minnella et al. 2007;
Tonello et al. 2007). Additionally, one
might speculate on a prolonged effect
after intravitreal injection as a result
of a decelerated release from the vitre-
ous.
So far, no data exist on the opti-
mum dosage to be applied in neovas-
cular glaucoma, nor on whether the
dosage should be adapted to the
degree of ischaemia. The intravitreal
dosage described in the literature
ranges from 1.0 mg (Kahook et al.
2006) to 1.5 mg (Silva Paula et al.
2006). The majority of surgeons
choose 1.25 mg bevacizumab based on
studies in neovascular age-related
macular degeneration (Avery 2006;
Davidorf et al. 2006; Iliev et al. 2006).
In this study, 1.25 mg bevacizumab
was injected intravitreally as well as
into the anterior chamber and an
acute decrease in IOP was achieved in
13 (65%) eyes with iris rubeosis after
intravitreal or intracameral treatment
with bevacizumab. According to the
results of this study, a dosage of
1.25 mg seems to be sufficient to
achieve a transient control of the
VEGF-driven progression of the
disease.
It is necessary to treat the underly-
ing angiogenic stimulus individually
for every patient, according to the
stage of disease and previous inter-
ventions. Three eyes with immediate
IOP control were treated only with
108
Fig. 5. Mechanisms of the different therapeutic strategies in neovascular glaucoma.
VEGF = vascular endothelial growth factor; IOP = intraocular pressure; AH = aqueous
humour.
bevacizumab injections and local anti-
glaucoma therapy. Nine eyes received
a combined therapy with at least one
further treatment. During the course
of disease, additional laser photocoag-
ulation to reduce VEGF secretion was
performed before or after intraocular
injection in 18 eyes and was the most
frequently applied additional treat-
ment. Filtering surgery or cyclode-
structive procedures were performed
in 16 (80%) eyes during the follow-up
period to reduce uncontrolled IOP. It
is likely that the adjunctive applica-
tion of bevacizumab in combination
with other treatment modalities in this
study extended the therapeutic win-
dow in such a way that the causative
treatment effects of, for example, pho-
tocoagulation, vitrectomy and periph-
eral retinal cryopexy had sufficient
time to make an impact. The combi-
nation therapy achieved a significant
longterm reduction in IOP and stabil-
ization in VA during the 12-month
follow-up. A reduction or even con-
clusion of local antiglaucoma therapy
was possible in 10 (50%) patients and
administration of acetazolamide was
not necessary in any patient at the
end of follow-up (100%). Intraocular
pressure control was achieved by the
combination of bevacizumab with
conventional therapeutic procedures
in all cases (100%).
Given its retrospective nature, this
study has several limitations, most of
all the lack of standardized guidelines
for a treatment protocol concerning a
stage-adapted combination with fur-
ther treatment modalities and the lack
of a control group. Nevertheless, this
study demonstrated that anti-angio-
genic therapy is an adjuvant approach
in neovascular glaucoma, which aims
to achieve an immediate reduction in
the pro-angiogenic factor VEGF as
the underlying stimulus and the asso-
ciated mechanisms such as vascular
permeability or vascular proliferation.
Therefore, anti-angiogenic therapy
should be an inherent part of any
stage-adapted treatment for neovascu-
lar glaucoma. For the therapeutic
approach, the stage of neovascular
disease should be considered. In early
rubeosis, an early application of bev-
acizumab may prevent the formation
of irreversible synechia. In advanced
disease bevacizumab may reduce
VEGF-controlled mechanisms such as
leakage or neovascularization and,
consequently, disease progression, but
it cannot reverse dense fibrovascular
membranes or contribute to an acute
IOP reduction. When planning ther-
apy, it should always be remembered
that anti-VEGF treatment does not
influence the cause of neovascular dis-
ease and relapse is not preventable as
long as the underlying stimulus (i.e.
an ischaemic retina) for VEGF pro-
duction is uninhibited by, for example,
panretinal photocoagulation. There-
fore, a causative treatment is essential
in every stage of disease. In cases of

persisting uncontrolled pressure, IOP
reduction may only be achieved by
additional cyclodestructive procedures
or filtering surgery (Fig. 5).
In conclusion, bevacizumab may be
utilized for short-term inhibition of
angiogenesis in neovascular glaucoma.
Early application of bevacizumab acts
as a preventative and may induce a
regression of early neovascular glau-
coma. In cases of advanced disease,
bevacizumab combined with conven-
tional therapeutic procedures is
required to reduce IOP. The causative
stimulus, such as an ischaemic retina,
must be treated in all cases.
References
Avery RL (2006): Regression of retinal and
iris neovascularization after intravitreal
bevacizumab (Avastin) treatment. Retina
26: 352–354.
Central Vein Occlusion Study Group (1995):
A randomized clinical trial of early panreti-
nal photocoagulation for ischaemic central
vein occlusion. Ophthalmology 102: 1434–
1444.
Davidorf FH, Mouser JG & Derick RJ
(2006): Rapid improvement of rubeosis
iridis from a single bevacizumab (Avastin)
injection. Retina 26: 354–356.
Diabetic Retinopathy Study Research Group
(1978): Photocoagulation treatment of pro-
liferative diabetic retinopathy: the second
report of Diabetic Retinopathy Study find-
ings. Ophthalmology 85: 82–106.
Eggleston TF, Bohling CA, Eggleston HC &
Hershey FB (1980): Photocoagulation for
ocular ischaemia associated with carotid
artery occlusion. Ann Ophthalmol 12: 84–
87.
Ehrenberg M, McCuen BW II, Schindler RH
& Machemer R (1984): Rubeosis iridis:
preoperative iris fluorescein angiography
and periocular steroids. Ophthalmology 91:
321–325.
Ferrara N, Hillan KJ, Gerber HP & Novotny
W (2004): Discovery and development of
bevacizumab, an anti-VEGF antibody for
treating cancer. Nat Rev Drug Discov 3:
391–400.
Grisanti S, Biester S, Peters S, Tatar O,
Ziemssen F, Bartz-Schmidt KU & the
Tübingen Bevacizumab Study Group
(2006): Intracameral bevacizumab for
iris rubeosis. Am J Ophthalmol 142: 158–
160.
Iliev ME, Domig D, Wolf-Schnurrbursch U,
Wolf S & Sarra GM (2006): Intravitreal
bevacizumab (Avastin) in the treatment of
neovascular glaucoma. Am J Ophthalmol
142: 1054–1066.
Kahook MY, Schuman JS & Noecker RJ
(2006): Intravitreal bevacizumab in a
patient with neovascular glaucoma. Oph-
thalmic Surg Lasers Imaging 37: 144–146.
Kernt M, Neubauer AS & Kampik A (2007):
Intravitreal bevacizumab (Avastin) treat-
ment is safe in terms of intraocular and
blood pressure. Acta Ophthalmol Scand
85: 119–120.
Mason JO III, Albert MA Jr, Mays A & Vail
R (2006): Regression of neovascular iris
vessels by intravitreal injection of bev-
acizumab. Retina 26: 839–841.
Acta Ophthalmologica 2010
Minnella AM, Savastano CM, Ziccardi L,
Scupola A, Sasso P, Falsini B & Balestrazzi
E (2007): Intravitreal bevacizumab (Ava-
stin) in proliferative diabetic retinopathy.
Acta Ophthalmol Scand [Epub ahead of
print].
Shima C, Sakaguchi H, Gomi F et al. (2008):
Complications in patients after intravitreal
injection of bevacizumab. Acta Ophthalmol
86: 372–376.
Silva Paula J, Jorge R, Alves Costa R, Rodri-
gues Mde L & Scott IU (2006): Short-term
results of intravitreal bevacizumab (Ava-
stin) on anterior segment neovasculariza-
tion in neovascular glaucoma. Acta
Ophthalmol Scand 84: 556–557.
Tonello M, Costa RA, Almeida FP, Barbosa
JC, Scott IU & Jorge R (2008): Panretinal
photocoagulation versus PRP plus intravi-
treal bevacizumab for high-risk prolifera-
tive diabetic retinopathy (IBeHi Study).
Acta Ophthalmol 86: 385–389.
Received on March 29th, 2008.
Accepted on May 12th, 2008.
Correspondence:
Prof. Dr Salvatore Grisanti
Department of Ophthalmology
University of Luebeck
Ratzeburger Allee 160
23538 Lübeck
Germany
Tel: +49 451 500 2211
Fax: +49 451 500 2671
Email: salvatore.grisanti@uk-sh.de
109
Copyright of Acta Ophthalmologica (1755375X) is the property of Blackwell Publishing Limited and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.