C L I N I C A L A N D E X P E R I M E N TA L
RESEARCH
Refractive error outcomes after intravitreal ranibizumab for
retinopathy of prematurity
Clin Exp Optom 2019
Qingyu Meng*†‡§ MD
Yong Cheng*†‡§ MD
Xi Wu*†‡§ MB
Dan Zhao¶ MB
Mingwei Zhao*†‡§ MD
Jianhong Liang*†‡§ MD
*Department of Ophthalmology and Clinical Centre of
Optometry, Peking University People’s Hospital,
Beijing, China
†
Eye Disease and Optometry Institute, Peking
University People’s Hospital, Beijing, China
‡
Beijing Key Laboratory of Diagnosis and Therapy of
Retinal and Choroid Diseases, Beijing, China
§ cent and 3.4 per cent at the three-month, six-month, 12-month and 24-month follow-up
College of Optometry, Peking University Health
Science Center, Beijing, China
¶
Department of Ophthalmology, The General Hospital
of Jixi Mining, Heilongjiang, China
E-mail: drliangjianhong@126.com
Submitted: 22 October 2018
Revised: 29 October 2019
Accepted for publication: 30 October 2019
Key words: anti-VEGF therapy, ranibizumab, refractive errors, retinopathy of prematurity
Retinopathy of prematurity (ROP) is a vascu-
lar proliferative disease caused by abnormal
development of retinal blood vessels in pre-
mature infants.1 ROP is the leading cause of
visual morbidity in children worldwide, and
is responsible for up to 60 per cent of visual
morbidity in middle-income countries.2
Ablation of the peripheral avascular retina
using cryotherapy and laser photocoagulation
has been the gold standard for reducing
the sequelae of ROP since the 1980s.3 In
the 1990s, vascular endothelial growth factor
(VEGF) was discovered to be related to retinal
neovascularisation. A prospective, controlled,
randomised, multicentre trial about anti-VEGF
treatment for ROP – Bevacizumab Eliminates
the Angiogenic Threat of Retinopathy of Pre-
maturity (BEAT-ROP) – showed that intravitreal
bevacizumab (IVB) was effective in treating
ROP and was more effective than laser
© 2019 Optometry Australia
DOI:10.1111/cxo.13019
Background: To evaluate the refractive outcomes in children treated with intravitreal injec-
tion of ranibizumab (IVR) for retinopathy of prematurity (ROP).
Methods: A retrospective study of 95 patients (186 eyes) was conducted. All patients
received IVR treatment. The cycloplegic refraction was evaluated at three months, six
months, 12 months and 24 months.
Result: The average spherical equivalent (SE) refraction value for patients with ROP who
were treated with IVR as a monotherapy at the three, six, 12 and 24 months was
+1.84  2.11 D, +1.02  2.41 D, +0.43  2.23 D and +0.13  2.73 D, respectively (p < 0.001).
Myopia (SE < −0.25 D) was observed in 24 eyes (15.9 per cent) at the three-month follow-up,
33 eyes (21.8 per cent) at the six-month follow-up, 33 eyes (26.5 per cent) at the 12-month
follow-up and 34 eyes (37.5 per cent) at the 24-month follow-up, respectively (p < 0.05). The
percentage of eyes with high myopia (SE < −5.0 D) was 0.6 per cent, 1.4 per cent, 1.7 per
visits in the IVR group. There were 59 eyes that received repeated IVR injections at the
follow-ups. The average SE of patients receiving repeated injections at the three-, six-, 12-
and 24-month visits was +1.53  2.03 D, +1.25  1.95 D, +0.58  2.24 D and −0.17  3.22 D,
respectively (p = 0.04).
Conclusion: Our large sample study found that 37.5 per cent and 3.4 per cent of patients
treated with IVR developed myopia and high myopia respectively, at the 24-month follow-
up. Furthermore, the mean SE decreased, and the trend of myopia increased, in the IVR
group at the follow-ups. Repeated injections might promote myopia in ROP patients. The
refractive status needs to be monitored in patients treated with ranibizumab.
treatment in patients with Zone I ROP (the cir- anterior chamber, and thickened crystalline
cle with a radius extending from the optic lens.9,10 These features are believed to be due
nerve to double the distance to the macula).4 to an arrested state of development in the
Ranibizumab is a monoclonal antibody frag- immature eye.11 Intravitreal anti-VEGF agents
ment (Fab) derived from the antibody of have the advantage of minimal influence on
bevacizumab but has approximately 10 times the local growth factor and signalling pathways
greater affinity for VEGF. Recently, a large ret- for the development of the anterior segment.
rospective study showed that intravitreal injec- The refractive outcomes after ranibizumab
tion of ranibizumab (IVR), as the primary treatment for ROP have rarely been
treatment, resulted in a positive response in reported.12,13 The aim of the current study was
94 per cent of ROP eyes.5 to evaluate the refractive changes in preterm
Several reports showed a lower prevalence infants with ROP two years after treatment
of myopia and high myopia in bevacizumab- with an intravitreal injection of ranibizumab.
treated patients than in laser-treated
patients.6–8 The BEAT-ROP clinical trial
reported that less myopia was found in eyes Methods
that received intravitreal bevacizumab than in
eyes that received laser treatment at age Participants
2.5 years.7 In all these cases, myopia was non- From April 2014 to March 2015, a retrospec-
axial with a steepened cornea, shallow tive study was performed in the Department
Clinical and Experimental Optometry 2019
1
Refractive error outcomes after intravitreal ranibizumab for ROP Meng, Cheng, Wu et al.

treatment.4 Zone 1 is the posterior zone of

Participants recruited
n = 186
Follow-up at 3 months
n = 152
Follow-up at 6 months
n = 142
Follow-up at 12 months
n = 117
Follow-up at 24 months
n = 88
Figure 1. Flowchart of patients included and follow-up
of Ophthalmology at Peking University Peo-
ple’s Hospital. Consecutive patients who
were diagnosed with ROP and accepted IVR
as the primary treatment were included.
This study was conducted with institutional
review board approval and adhered to the
tenets of the Declaration of Helsinki.
According to Chinese guidelines, infants
who were born at a gestational age of
< 32 weeks and whose birth weight was
< 2,000 g were screened for ROP.14 The first
screening was at four to six weeks after
birth or at 31 to 32 weeks postmenstrual
age (PMA), whichever came first.14 ROP
25 were ROP 4B or worse
7 with glaucoma
5 with cornea opacity
5 with congenital cataract
5 progressed to ROP 4B
5 lost to follow-up
5 with glaucoma
20 lost to follow-up
29 lost to follow-up
stage was determined according to the
International Classification of Retinopathy of
Prematurity 2005.15 At the screening and
follow-ups, all patients had fundus photog-
raphy by the RetCam Imaging System
(Clarity Medical System, Pleasanton, CA,
USA). The cycloplegic agent used was 0.5%
tropicamide/0.5% phenylephrine eye drops.
Treatment and follow-up
Infants who were diagnosed with type
1 ROP, as defined by the Early Treatment for
Retinopathy of Prematurity (ETROP) criteria,
were treated with IVR as the initial
the retina, defined as the circle with a radius
extending from the optic nerve to double the
distance to the macula. Zone 2 is an annulus
with the inner border defined by Zone 1 and
the outer border defined by the radius
defined as the distance from the optic nerve
to the nasal ora serrata. Zone 3 is the resid-
ual temporal crescent of the retina. Zone
1 with any stage with plus disease, Zone
1 with stage 3 without plus disease, and
Zone 2 with stage 2 or 3 with plus disease, or
Zone 2, with more than five contiguous or
eight cumulative clock hours were treated.3
A 0.25 mg/0.025 ml dose of ranibizumab
(half of the dose administered intravitreally
in adults for ocular diseases) was injected
into each eye by experienced surgeons.5 The
patients were re-examined the next day and
then every week thereafter to monitor the
progression of the disease until full
vascularisation was noted or additional treat-
ment was given. Recurrence was defined as
the reappearance of disease with or without
plus disease after full regression had been
obtained by primary treatment. Infants with
recurrence were followed more closely. In
patients with ROP recurrence, additional
treatment including repeated injections, laser
photocoagulation or surgery was applied.
Additional treatments were not considered if
peripheral avascular areas remained stable
during the follow-up. Patients who were not
able to undergo the examinations, who
developed any form of cataracts or received
cataract surgery, who developed corneal
opacity, who had stage 4B ROP or worse, or
who developed glaucoma were excluded.
The eyes of the infants were assessed to
identify refractive errors 30 minutes after

Follow-up (months) 3 6 12 24
Eyes (n) 152 142 117 88
Gestational age (weeks, means  SD) 28.9  1.8 29.0  1.8 29.1  1.8 28.9  1.8
Birth weight (g, means  SD) 1,270.5  373.3 1,237.9  307.6 1,271.2  319.7 1,243.1  324.6
APROP (n) 28 28 23 20
Zone 1 (n) 5 9 6 4
Zone 2 (n) 119 105 89 64
Numbers with repeated injections (n, %) 59 (38.8) 57 (40.1) 48 (41.0) 43 (48.9)
Numbers with additional laser treatment (n, %) 11 (7.2) 8 (5.6) 7 (5.9) 5 (5.7)
APROP: aggressive posterior retinopathy of prematurity, ROP: retinopathy of prematurity, SD: standard deviation.
Zone 1 is the circle with a radius extending from the optic nerve to double the distance to the macula.
Zone 2 is an annulus with the inner border defined by Zone 1 and the outer border defined by the radius defined as the distance
from the optic nerve to the nasal ora serrata.

Table 1. Demographic information for the treated ROP patients

Clinical and Experimental Optometry 2019 © 2019 Optometry Australia

2
 Refractive error outcomes after intravitreal ranibizumab for ROP Meng, Cheng, Wu et al.

Follow-up (months) 3 (n = 152) 6 (n = 142) 12 (n = 117) 24 (n = 88) p-value OR (95% CI)

Myopia (n, %) 24 (15.9) 33 (21.8) 33 (26.5) 34 (37.5) 0.03 –
High myopia (n, %) 1 (0.6) 2 (1.4) 2 (1.7) 3 (3.4) 0.46 –
High astigmatism 17 (11.2) 19 (13.4) 12 (10.3) 2 (2.3) 0.08 –
SE (D) +1.88  2.17 +1.11  2.33 +0.55  2.05 +0.20  2.68 0.00 −1.92 (−0.210, −1.65)
Spherical (D) +1.49  1.77 +0.97  1.94 +0.61  1.65 +0.33  2.24 0.00 0.12 (−0.08, 0.32)
J0 −0.05  0.53 +0.64  0.56 −0.01  0.47 −0.02  0.43 0.51 0.15 (−0.11, 0.41)
J45 +0.02  0.53 −0.06  0.56 −0.01  0.53 +0.02  0.46 0.72 −0.12 (−0.28, 0.06)
CI: confidence interval, OR: odds ratio, SE: spherical equivalent.

Table 2. Refractive error results of patients

three measurements were taken to obtain

2.50 an average reading. Myopia was defined as
an SE ≤ −0.25 D, and high myopia was
defined as an SE ≤ −5.00 D. High astigma-
2.00 tism was defined as a minus cylinder form
> 1.50 D. Emmetropia was defined as an
SE > −0.25 D but < 2 D.
1.50
Mean SE

Sample size calculation

1.00 Sample size was calculated using the follow-
ing equation: n = (U1-α/2*σ/δ)2, n being the
sample size, σ is the standard deviation; δ is
0.50 the anticipation error; α = 0.05. According to
a previous study,12 the standard deviation
of SE after IVR treatment was +0.48. Antici-
0.00
pation error was 0.1. The result of the calcu-
lation showed that the minimum sample
-0.50 size required was 88 eyes.

3 6 12 24 Power vector analysis

To better describe the optical characteristics,
Month we used the method of power vector analy-
sis. As described in previous reports,16,17 we
Figure 2. Mean spherical equivalent (SE) results in the intravitreal injection of
converted the refractive astigmatism from
ranibizumab (IVR) group. (The error bars represent standard deviation.)
the spherocylindrical notation to the power
vector notation. J0 represents the power of
instilling 0.5% tropicamide/0.5% phenyleph- refraction. The results were recorded in the the orthogonal astigmatism. Positive values
rine (two instillations at an interval of form of spherical and cylinder power, astig- of J0 indicate with-the-rule astigmatism,
10 minutes). A handheld autorefractometer matism degree and axis, type of astigma- and negative values indicate against-the-
(FR-5000; Grand Seiko Co., Ltd., Hiroshima, tism, and spherical equivalent (SE, spherical rule astigmatism. J45 represents oblique
Japan) was used to evaluate the cycloplegic plus half of the cylinder power). At least astigmatism.

Follow-up (months) 3 (n = 59) 6 (n = 57) 12 (n = 48) 24 (n = 43) p-value OR (95% CI)

Myopia (n, %) 9 (15.3) 13 (22.8) 15 (31.3) 17 (39.5) 0.03
High myopia (n, %) 1 (1.7) 0 (0.0) 2 (4.2) 4 (9.3) 0.06
High astigmatism (n, %) 22 (37.3) 22 (38.6) 15 (31.3) 12 (27.9) 0.64
SE (D) +1.96  2.03 +1.25  1.95 +0.58  2.24 −0.17  3.22 0.04 0.49 (0.25, 0.73)
Spherical (D) +1.53  1.75 +1.11  1.48 +0.60  1.81 −0.07  2.71 0.04 0.25 (0.02, 0.48)
J0 +0.34  0.63 +0.22  0.58 −0.1  0.50 −0.07  0.43 0.85 −0.56 (−0.55, 1.25)
J45 +0.11  0.53 −0.02  0.49 −0.00  0.53 +0.03  0.52 0.72 0.12 (−0.24, 0.46)
CI: confidence interval, OR: odds ratio, SE: spherical equivalent.

Table 3. Refractive error results of patients who received repeated intravitreal injection of ranibizumab

© 2019 Optometry Australia Clinical and Experimental Optometry 2019

3
Refractive error outcomes after intravitreal ranibizumab for ROP Meng, Cheng, Wu et al.

J0 was calculated by using the following cent) at the six-month follow-up, 33 eyes
equation: J0 = (-C/2)*cos2A. (26.5 per cent) at the 12-month follow-up and

(−0.06, 0.18)

(−0.36, 0.15)

(−0.24, 0.22)
p-value OR (95% CI)

(0.74, 1.43)

(1.00, 1.21)
−0.02

−0.01
J45 was calculated by using the following 34 eyes (37.5 per cent) at the 24-month

−0.1
1.03

1.10

equation: J45 = (-C/2)*sin2A. follow-up (p < 0.05). One eye (0.6 per cent),
C is the negative cylinder power, and A is two eyes (1.4 per cent), two eyes (1.7 per
0.88

0.03

0.27

0.42

0.93
the cylinder axis. cent) and three eyes (3.4 per cent) at the
three-month, six-month, 12-month and
−0.072  0.43

Data analysis
−0.17  3.22

+0.03  0.52

24-month follow-ups were discovered to have

injections
Repeated
24

17 (39.5)

4 (9.3)

We performed statistical analysis with the high myopia (SE < −5.00 D) (p = 0.46). High
program IBM SPSS 22 (IBM, Armonk, NY, astigmatism was observed in 17 eyes (11.2
USA). Chi-squared and Fisher’s exact tests per cent) at the three-month follow-up,
p-value OR (95% CI) Monotherapy

−0.03  0.39
+0.59  1.98

+0.03  0.43

were performed to compare categorical 19 eyes (13.4 per cent) at the six-month
17 (37.8)

0 (0.0)

data. Correlation test was performed to follow-up, 12 eyes (10.3 per cent) at the
evaluate if there was correlation in refrac- 12-month follow-up and two eyes (2.3 per
tive error between the two eyes in each par- cent) at the 24-month follow-up (p = 0.08).
(−0.04, 0.04)

(−0.14, 0.05)
(−0.2, 0.19)
(0.79, 1.29)

(0.98, 1.11)

Table 4. Comparison of refractive errors between patients with IVR monotherapy and those with repeated IVR injections

ticipant. In that case, mixed model testing

−0.006
0.002

0.35
1.01

1.04

was performed to compare the means of Refractive error outcome of

refractive error. Otherwise, the Kruskal- patients receiving repeated
Wallis test was used to compare the means. injections
0.92

0.09

0.94

0.95

0.70

A p-value less than 0.05 was considered sig- There were 59 eyes that received repeated
nificant for all tests. IVR injections during the follow-ups. The
−0.02  0.54 −0.00  0.53
+0.58  2.24

−0.1  0.50
injections
Repeated
12

15 (31.3)

refractive error outcomes are shown in

2 (4.2)

Table 3. The mean SE decreased as time

Results progressed in patients with repeated injec-
p-value OR (95% CI) Monotherapy

−0.01  0.47
+0.52  1.91

tions (p = 0.04). Myopia was observed in

21 (30.4)

0 (0.0)

Patient characteristics nine eyes (15.3 per cent) at the three-month

This study consisted of 186 eyes of follow-up, 13 eyes (22.8 per cent) at the six-
95 patients, all of whom received intravitreal month follow-up, 15 eyes (31.3 per cent) at
(−0.02, 0.05)

(−0.23, 0.07)

(−0.08, 0.22)
(0.83, 1.19)

(0.93, 1.01)

ranibizumab (IVR group). In the IVR group, the 12-month follow-up and 17 eyes (39.5
−0.08
0.99

0.03
0.99

0.07

the mean gestational age and birth weight
of the patients were 29.1  1.8 weeks (range
26.2 to 33.6 weeks) and 1,260.3  351.5 g
CI: confidence interval, IVR: intravitreal injection of ranibizumab, OR: odds ratio, SE: spherical equivalent.
0.98
per cent) at the 24-month follow-up
(p < 0.05). One eye (1.7 per cent), two eyes
(4.2 per cent), and four eyes (9.3 per cent) at

0.16

0.37

0.37

0.31

(range 780 to 3,000 g), respectively. There the three-month, 12-month and 24-month
were 152, 142, 117 and 88 eyes assessed to follow-ups, respectively, were discovered to
−0.08  0.61 −0.02  0.49
+1.25  1.95

+0.22  0.58
injections
Repeated

13 (22.8)
6

identify refractive errors at the three-month, have high myopia (p = 0.06).

0 (0.0)

six-month, 12-month and 24-month follow- We compared the refractive error

up visits, respectively. The patient character- between patients with IVR monotherapy
p-value OR (95% CI) Monotherapy

istics are shown in Figure 1. The mean and those with repeated IVR injections
+1.02  2.57

+0.09  0.55
20 (22.9)

3 (3.4)

gestational age of patients at the three- (Table 4). There were no significant differ-
month, six-month, 12-month and 24-month ences in SE between patients with IVR mon-
follow-up visits was 28.9  1.8 weeks, 29.0  otherapy and repeated injections (p > 0.05).
(−0.37, 0.04)

(−0.08, 0.21)

(−0.21, 0.27)

1.8 weeks, 29.1  1.8 weeks and 28.9  At the 24-month follow-up, ROP patients
(0.86, 1.14)

(0.98, 1.05)
0.001
1.02

0.13
0.99

0.06

1.8 weeks, respectively. The demographic who received repeated injections had higher
information for all patients is presented in percentages of high myopia than patients
Table 1. with IVR monotherapy (p = 0.03).
0.093
0.88

0.21

0.96

0.39

Refractive error outcomes at Refractive error outcome of

+1.96  2.03

−0.04  0.47 +0.34  0.63

−0.04  0.52 +0.11  0.53

injections
Repeated

follow-up patients with additional laser

3

9 (15.3)

1 (1.7)

The refractive outcomes after intravitreal treatment

ranibizumab treatment are shown in Table 2. There were 11 eyes that received additional
Monotherapy

The average SE of patients at the three- laser treatment during follow-up. At

+1.87  2.07
15 (16.1)

month, six-month, 12-month and 24-month the three-month, six-month, 12-month and
0 (0.0)

follow-up visits was +1.88  2.17 D, +1.11  24-month follow-ups, the average SE was
2.33 D, +0.55  2.05 D and + 0.20  2.68 D, +1.92  2.18 D, +1.25  3.06 D, −0.82 
4.14 D and −2.05  2.87 D, respectively
Myopia (n, %)

respectively (p < 0.001). The mean SE is pro-

High myopia
Follow-up
(months)

vided in Figure 2. The mean SE decreased as (p = 0.023). The data are shown in Table 5.
SE (D)

time progressed. Myopia (SE < −0.25 D) was

(n, %)

The percentage of eyes with myopia or with

J45
J0

observed in 24 eyes (15.9 per cent) at the high myopia showed no significant differ-
three-month follow-up, 33 eyes (21.8 per ence at all follow-ups (p > 0.1).

Clinical and Experimental Optometry 2019 © 2019 Optometry Australia

4
 Refractive error outcomes after intravitreal ranibizumab for ROP Meng, Cheng, Wu et al.

Follow-up (months) 3 (n = 11) 6 (n = 8) 12 (n = 7) 24 (n = 5) p-value

Gestational age 29.1  1.5 29.3  1.6 29.9  0.9 30.0  1.1 –
(weeks, means  SD)
Birth weight 1,270.5  1,237.9  1,271.2  1,243.1  –
(g, means  SD) 373.3 307.6 319.7 324.6
Myopia (n, %) 1 (9.1) 2 (25.0) 2 (28.6) 4 (80.0) 0.27
High myopia (n, %) 0 (0.0) 0 (0.0) 2 (28.6) 1 (20.0) 0.23
High astigmatism (n, %) 4 (36.4) 1 (12.5) 2 (28.6) 0 (0.0) 0.53
SE (D) +1.92  2.18 +1.25  3.06 −0.82  4.14 −2.05  2.87 0.03
Spherical (D) +1.18  2.16 +0.46  3.16 −0.67  3.4 −1.85  2.52 0.09
J0 0.34  0.63 0.22  0.58 −0.1  0.50 −0.072  0.43 0.34
J45 0.11  0.53 −0.02  0.49 −0.00  0.53 0.03  0.52 0.43
SD: standard deviation, SE: spherical equivalent.

Table 5. Refractive error results of patients with additional laser treatment

Author Publication Country Number Treatment Mean age at Mean SE Myopia (%) High
date of eyes examination myopia (%)
Harder6 2012 Germany 12 IVB 0.88 0.64
Martinez-Castellanos26 2013 Mexico 9 IVB 2 −1.0 55.6 11.1
Harder8 2013 Germany 23 IVB 1 −1.04 NA 17.0
Chen21 2014 Taiwan 40 IVB 2 −0.98 47.5 10.0
Chen21 2014 Taiwan 17 IVB + laser 2 −2.4 82.4 29.4
Geloneck7 2014 USA 52 IVB 2.5 −1.51 NA 3.8
Hwang27 2015 Georgia 22 IVB 1.87 −2.4 NA NA
Kabataş13 2017 Turkey 24 IVB 1.5 −1.49 NA NA
Kabataş13 2017 Turkey 12 IVR 1.5 −1.79 NA NA
Chen12 2015 Taiwan 41 IVB 1 −0.3 NA 14.6
Chen12 2015 Taiwan 31 IVR 1 0.1 NA 0
Current study 2017 China 88 IVR 2 0.20 34.0 3.0
Current study 2017 China 5 IVR + laser 2 −2.05 80.0 20
IVB: intravitreal injection of bevacizumab, IVR: intravitreal injection of ranibizumab, SE: spherical equivalent, VEGF: vascular endothe-
lial growth factor.

Table 6. Refractive outcomes after anti-VEGF treatment for retinopathy of prematurity

Discussion
Current estimates suggest that of the 13 mil-
lion premature children who are born
worldwide each year who survive the neo-
natal period, vision-threatening ROP will
develop in more than 50,000 of them.18
However, the visual rehabilitation of these
children does not end with ROP regression.
One important problem frequently found to
affect these children is the development of
myopia, especially very high myopia, which
has considerable structural and functional
consequences. The intravitreal injection of
anti-VEGF agents offers a promising option
in the treatment armamentarium with the
advantage of being less time consuming
than laser treatment, potentially less risky
by avoiding general anaesthesia, and being
able to allow further retinal vascularisation
in treatment-requiring cases.19
Lin et al. showed that intravitreal injection
of ranibizumab seemed to be effective in
treating patients with ROP, with 94 per cent
showing a positive response, and some
authors recommending IVR instead of IVB to
treat severe ROP.20 The reported rates of
myopia and high myopia in participants
who received IVB ranged from 47.5 per cent
to 55.6 per cent and 10.0 per cent to 17.0
per cent, respectively (Table 6). Chen et al.21
reported a higher prevalence of myopia and
high myopia in the IVB + laser group than in
the IVB group. Previously, there have only
been two small sample size studies that
evaluated the refractive error outcomes
after IVR therapy for ROP.12,13 The reported
mean SE was between −1.79 D and +0.1 D
for patients between 1 and 1.5 years old,
and these results are similar to that of our
patients who were treated with IVR only
(+0.2 D). Chen et al.12 reported that the
prevalence of high myopia was 14.6 per
cent in the bevacizumab group and zero per
cent in the ranibizumab group, that is, the
bevacizumab group had a significantly higher
prevalence of high myopia than the
ranibizumab group (p = 0.03). Kabataş et al.13
observed myopic SE in the IVB group and the
IVR groups but without any significant differ-
ence in Turkish ROP patients. In the current
study, the percentages of patients with myo-
pia and high myopia among IVR-treated

© 2019 Optometry Australia Clinical and Experimental Optometry 2019

5
Refractive error outcomes after intravitreal ranibizumab for ROP Meng, Cheng, Wu et al.
patients were 34.0 per cent and three per
cent, respectively, and these values are lower
than the previously reported percentages for
patients treated with IVB. Furthermore, we
also found that the mean SE decreased as
time progressed in patients with repeated
injections. In total 17 eyes (39.5 per cent) had
myopia and four eyes (9.3 per cent) had high
myopia at the 24-month follow-up in the
repeated IVR group. Compared to the IVR
monotherapy group, repeated injections had
higher percentages of high myopia (0.0 per
cent versus 9.3 per cent, p = 0.03) at the
24-month follow-up. Our results also
suggested that repeated injections could pro-
mote myopia in ROP patients.
Researchers have hypothesised that the dif-
ference in prevalence of high myopia in IVB-
treated and IVR-treated patients may be due to
differences in the regulation of retinal VEGF
between bevacizumab and ranibizumab.12
Because bevacizumab has a longer half-life
than ranibizumab, the difference in prevalence
may be caused by sustained inhibition of VEGF
by bevacizumab, which may induce more apo-
ptosis in the peripheral retina, with subsequent
dysregulation of emmetropisation.
Our study is limited by the retrospective
study design, retinoscopy not being per-
formed in any of the patients, and the lack of
biometric measurements. Moreover, in our
study, the percentage of patients lost to
follow-up was 42.1 per cent. Among these par-
ticipants, 15.6 per cent progressed to ROP 4B
or worse, or had glaucoma, while other partici-
pants did not return to the clinic. The reason
may be that nearly half of the participants
came from rural areas. These participants
tend to have their eyes re-examined in the
hospitals near their homes. Another limitation
is that 0.5% tropicamide/0.5% phenylephrine
eye drops was used as the cycloplegic agent
in our study. Tropicamide is a convenient
agent used in China because it requires
shorter periods of time to obtain maximum
cycloplegia and to recover from cycloplegia,
and tropicamide rarely induces systemic side
effects, including blurred vision and photopho-
bia. Several early studies indicated that the
cycloplegic effect of tropicamide was much
less than that of cyclopentolate or atropine
and was thus insufficient for refractive mea-
surements.22,23 However, recent studies have
proposed different results. The Correction of
Myopia Evaluation Trial (COMET) concluded
that 1% tropicamide is an effective cycloplegic
agent in myopic children24 and Hamasaki
et al. have suggested that eye drops that are
a mixture of 0.5% tropicamide and 0.5%
Clinical and Experimental Optometry 2019
phenylephrine are an acceptable and useful
cycloplegic agent in Japan.25 As Japanese peo-
ple have the same dark iris colour as Chinese
people, we believe their findings could be
safely applied to our study. Another limitation
is the small number of patients in the IVR +
laser group, leading to the insignificant differ-
ence in refractive error results in the two
groups (data not shown). Despite these limita-
tions, this study may still provide useful infor-
mation for ROP patients who were treated
with IVR therapy. Future prospective studies
with large sample sizes are warranted to eluci-
date whether there are any differences in
treatment outcomes between these two
medications.
In conclusion, our large sample study
found that 37.5 per cent and 3.4 per cent of
patients treated with IVR developed myopia
and high myopia, respectively, at the
24-month follow-up. Furthermore, the mean
SE decreased and the trend of myopia
increased in the IVR group at the follow-ups.
Refractive status needs to be monitored in
patients treated with ranibizumab.
ACKNOWLEDGEMENTS
The present study was supported by Peking
University People’s Hospital Research and
Development Funds (grant no. RDY-
2017-2017). We are grateful for the assis-
tance provided by the Scientific Research
Department of Peking University People’s
Hospital in data analysis.
REFERENCES
1. Section on Ophthalmology; American Academy of
Pediatrics; American Academy of Ophthalmology and;
American Association for Pediatric Ophthalmology
and Strabismus. Screening examination of premature
infants for retinopathy of prematurity. Pediatrics 2006;
117: 572–576.
2. Gilbert C, Fielder A, Gordillo L et al. Characteristics of
infants with severe retinopathy of prematurity in
countries with low, moderate, and high levels of
development: implications for screening programs.
Pediatrics 2005; 115: e518–e525.
3. Early Treatment for Retinopathy of Prematurity Coop-
erative G. Revised indications for the treatment of ret-
inopathy of prematurity: results of the early
treatment for retinopathy of prematurity randomized
trial. Arch Ophthalmol 2003; 121: 1684–1694.
4. Mintz-Hittner HA, Kennedy KA, Chuang AZ et al. Effi-
cacy of intravitreal bevacizumab for stage 3+ retinopa-
thy of prematurity. N Engl J Med 2011; 364: 603–615.
5. Huang Q, Zhang Q, Fei P et al. Ranibizumab injection
as primary treatment in patients with retinopathy of
prematurity: anatomic outcomes and influencing fac-
tors. Ophthalmology 2017; 124: 1156–1164.
6. Harder BC, von Baltz S, Schlichtenbrede FC et al. Early
refractive outcome after intravitreous bevacizumab
for retinopathy of prematurity. Arch Ophthalmol 2012;
130: 800–801.
7. Geloneck MM, Chuang AZ, Clark WL et al. Refractive
outcomes following bevacizumab monotherapy
compared with conventional laser treatment: a
randomized clinical trial. JAMA Ophthalmol 2014; 132:
1327–1333.
8. Harder BC, Schlichtenbrede FC, von Baltz S et al. Intra-
vitreal bevacizumab for retinopathy of prematurity:
refractive error results. Am J Ophthalmol 2013; 155:
1119–1124, e1111.
9. Hittner HM, Rhodes LM, McPherson AR. Anterior seg-
ment abnormalities in cicatricial retinopathy of pre-
maturity. Ophthalmology 1979; 86: 803–816.
10. Chen TC, Tsai TH, Shih YF et al. Long-term evaluation
of refractive status and optical components in eyes of
children born prematurely. Invest Ophthalmol Vis Sci
2010; 51: 6140–6148.
11. Wang J, Ren X, Shen L et al. Development of refractive
error in individual children with regressed retinopathy
of prematurity. Invest Ophthalmol Vis Sci 2013; 54:
6018–6024.
12. Chen SN, Lian I, Hwang YC et al. Intravitreal anti-vascular
endothelial growth factor treatment for retinopathy of
prematurity: comparison between Ranibizumab and
Bevacizumab. Retina 2015; 35: 667–674.
13. Kabatas EU, Kurtul BE, Altiaylik Ozer P et al. Compari-
son of Intravitreal Bevacizumab, Intravitreal
Ranibizumab and laser photocoagulation for treat-
ment of type 1 retinopathy of prematurity in Turkish
preterm children. Curr Eye Res 2017; 42: 1054–1058.
14. Chen Y, Feng J, Gilbert C et al. Time at treatment of
severe retinopathy of prematurity in China: recom-
mendations for guidelines in more mature infants.
PLoS One 2015; 10: e0116669.
15. International Committee for the Classification of Reti-
nopathy of P. The international classification of reti-
nopathy of prematurity revisited. Arch Ophthalmol
2005; 123: 991–999.
16. Namba H, Sugano A, Nishi K et al. Age-related varia-
tions in corneal geometry and their association with
astigmatism: the Yamagata study (Funagata). Medicine
(Baltimore) 2018; 97: e12894.
17. Thibos LN, Wheeler W, Horner D. Power vectors: an
application of Fourier analysis to the description and
statistical analysis of refractive error. Optom Vis Sci
1997; 74: 367–375.
18. Blencowe H, Lawn JE, Vazquez T et al. Preterm-
associated visual impairment and estimates of reti-
nopathy of prematurity at regional and global levels
for 2010. Pediatr Res 2013; 74: 35–49.
19. VanderVeen DK, Melia M, Yang MB et al. Anti-vascular
endothelial growth factor therapy for primary treat-
ment of type 1 retinopathy of prematurity: a report
by the American Academy of ophthalmology. Ophthal-
mology 2017; 124: 619–633.
20. Lin CJ, Chen SN, Hwang JF. Intravitreal ranibizumab as
salvage therapy in an extremely low-birth-weight
infant with rush type retinopathy of prematurity.
Oman J Ophthalmol 2012; 5: 184–186.
21. Chen YH, Chen SN, Lien RI et al. Refractive errors after
the use of bevacizumab for the treatment of retinopa-
thy of prematurity: 2-year outcomes. Eye (Lond) 2014;
28: 1080–1086.
22. Milder B. Tropicamide as a cycloplegic agent. Arch
Ophthalmol 1961; 66: 70–72.
23. Lovasik JV. Pharmacokinetics of topically applied
cyclopentolate HCl and tropicamide. Am J Optom Phy-
siol Opt 1986; 63: 787–803.
24. Manny RE, Hussein M, Scheiman M et al. Tropicamide
(1%): an effective cycloplegic agent for myopic chil-
dren. Invest Ophthalmol Vis Sci 2001; 42: 1728–1735.
25. Hamasaki I, Hasebe S, Kimura S et al. Cycloplegic effect
of 0.5% tropicamide and 0.5% phenylephrine mixed eye
drops: objective assessment in Japanese schoolchildren
with myopia. Jpn J Ophthalmol 2007; 51: 111–115.
26. Martinez-Castellanos MA, Schwartz S, Hernandez-
Rojas ML et al. Long-term effect of antiangiogenic
therapy for retinopathy of prematurity up to 5 years
of follow-up. Retina 2013; 33: 329–338.
27. Hwang CK, Hubbard GB, Hutchinson AK et al. Out-
comes after Intravitreal Bevacizumab versus laser
photocoagulation for retinopathy of prematurity: a
5-year retrospective analysis. Ophthalmology 2015;
122: 1008–1015.
© 2019 Optometry Australia