Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns,

and Perspectives
Luca Filippi, MD1, Giacomo Cavallaro, MD2, Paola Bagnoli, PhD3, Massimo Dal Monte, PhD3, Patrizio Fiorini, MD1,
Gianpaolo Donzelli, MD1, Francesca Tipelli, MD4, Gabriella Araimo, MD2, Gloria Cristofori, MD2, Giancarlo la Marca, Pharm Sc5,
Maria Luisa Della Bona, Pharm Sc5, Agostino La Torre, MD6, Pina Fortunato, MD6, Sandra Furlanetto, Pharm Sc7,
Silvia Osnaghi, MD8, and Fabio Mosca, MD2

Objective To evaluate safety and efficacy of oral propranolol administration in preterm newborns affected by an
early phase of retinopathy of prematurity (ROP).
Study design Fifty-two preterm newborns with Stage 2 ROP were randomized to receive oral propranolol (0.25 or
0.5 mg/kg/6 hours) added to standard treatment or standard treatment alone. To evaluate safety of the treatment,
hemodynamic and respiratory variables were continuously monitored, and blood samples were collected weekly to
check for renal, liver, and metabolic balance. To evaluate efficacy of the treatment, the progression of the disease
(number of laser treatments, number of bevacizumab treatments, and incidence of retinal detachment) was evalu-
ated by serial ophthalmologic examinations, and plasma soluble E-selectin levels were measured weekly.
Results Newborns treated with propranolol showed less progression to Stage 3 (risk ratio 0.52; 95% CI 0.47-0.58,
relative reduction of risk 48%) or Stage 3 plus (relative risk 0.42 95% CI 0.31-0.58, relative reduction of risk 58%).
The infants required fewer laser treatments and less need for rescue treatment with intravitreal bevacizumab (rela-
tive risk 0.48; 95% CI 0.29-0.79, relative reduction of risk 52 %), a 100% relative reduction of risk for progression to
Stage 4. They also had significantly lower plasma soluble E-selectin levels. However, 5 of the 26 newborns treated
with propranolol had serious adverse effects (hypotension, bradycardia), in conjunction with episodes of sepsis,
anesthesia induction, or tracheal stimulation.
Conclusion This pilot study suggests that the administration of oral propranolol is effective in counteracting the
progression of ROP but that safety is a concern. (J Pediatr 2013;-:---).

D
espite progressive improvements in neonatal care, retinopathy of prematurity (ROP) remains the major cause of blind-
ness and visual impairment in children in both developing and industrialized countries.1 The incidence of the disease is
closely related to birth weight and gestational age (GA): ROP is more severe and more frequent in extremely premature
infants and in newborns with extremely low birth weights.2
The pathogenesis of ROP is hypothesized to consist of 2 distinct phases, of which the second phase is characterized by
hypoxia-induced up-regulation of vascular endothelial growth factor (VEGF) and retinal neovascularization.3 As understand-
ing of the pathophysiology of ROP has increased, emphasis has shifted to selective therapies that target components of the
angiogenesis cascade. There are some indications that the b-adrenergic system may interfere with ROP in infants. For instance,
beta-adrenergic receptor (b-AR) polymorphisms existing in many black infants4 may be responsible for the lower incidence of
ROP progression in these infants compared with nonblack infants.2,5 In fact, the polymorphism of G-protein coupled recep-
tor kinase (ie, GRK5-L41) facilitates b-AR phosphorylation, recruitment of b-ar-
restin, and consequent desensitization during catecholamine excess, making this
population genetically b-blocked.6 In addition, b-AR blockade with propranolol, 1
From the Neonatal Intensive Care Unit, Medical Surgical
Fetal-Neonatal Department, “A. Meyer” University
a nonselective b1-AR and b2-AR blocker,7 induced involution of infantile hem- 2
Children’s Hospital, Florence; Neonatal Intensive Care
angioma,8 the most common tumor of infancy that is often associated with Unit, Fondazione Istituto di Ricovero e Cura a Carattere
Scientifico Ca
Granda Ospedale Maggiore Policlinico,
ROP,9 suggesting that b-AR blockers might be effective in ROP as well. Universita
degli Studi di Milano, Milan; Department of
Biology, Unit of General Physiology, University of Pisa;
3

These observations are supported by experimental findings in a mouse model 4

Department of Developmental Neuroscience, IRCCS
5
Stella Maris, Calambrone, Pisa; Department of Pediatric
of oxygen-induced retinopathy (OIR),10,11 where we demonstrated that retinal Neurosciences, “A. Meyer” University Children’s
norepinephrine increases in response to hypoxia and b-ARs regulate VEGF pro-
6
Hospital; Pediatric Ophthalmology, Careggi University
7
Hospital; Department of Pharmaceutical Sciences,
12-14
duction and retinal neovascularization. Most interesting, these studies also University of Florence, Florence; and Department of 8

Ophthalmology, Fondazione IRCCS Ca
Granda,

degli Studi di
Ospedale Maggiore Policlinico, Universita
Milano, Milan, Italy
The authors declare no conflicts of interest.
b-AR b-adrenoreceptor Registered with ClinicalTrials.gov (NCT01079715), Cur-
GA Gestational age rent Controlled Trials (ISRCTN18523491), and European
OIR Oxygen-induced retinopathy Union Drug Regulating Authorities Clinical Trials (2010-
018737-21).
ROP Retinopathy of prematurity
VEGF Vascular endothelial growth factor 0022-3476/$ - see front matter. Copyright ª 2013 Mosby Inc.
All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.07.049

1
THE JOURNAL OF PEDIATRICS  www.jpeds.com
showed that significant reduction of neovascularization
could be achieved by the administration of topical propran-
olol,15 suggesting the therapeutic use of b-AR blockers to
counteract retinal neovascularization in ROP.
Assuming in human preterm newborns with ROP that
VEGF overexpression and retinal neovascularization in
response to hypoxia might involve b-AR activation, we de-
signed a randomized pilot trial to test whether the oral admin-
istration of propranolol in newborns with a precocious stage
of ROP is safe and might reduce the progression of the disease.
Methods
We used the standard international classification to establish
the location and severity of ROP.16 A randomized controlled
pilot trial was performed with preterm newborns with GA
<32 weeks of age and Stage 2 ROP without plus in zone II
who were admitted to 2 neonatal intensive care units, both
regional reference centers for neonatal surgical diseases.
Although infants were receiving supplemental oxygen, the
target range of oxygen saturation was maintained between
91% and 95%. We randomized treated and control newborns
with a 1:1 allocation in blocks of 8 by using a computer
random number generator and stratified by center in 2
groups of GA 23-25 and 26-32 weeks; they alternated between
propranolol added to standard treatment or standard treat-
ment alone (the treatment adopted by the Early Treatment
of Retinopathy of Prematurity Cooperative Group).17 The
allocation sequence was concealed in sequentially numbered,
opaque, sealed envelopes. Exclusion criteria included new-
borns with congenital or acquired cardiovascular anomalies,
renal failure or cerebral hemorrhage at enrollment, and new-
borns with ROP in zone I or at a more advanced stage than
Stage 2 without plus in zone II. With severe adverse effects
related to propranolol (severe bradycardia, hypotension, or
wheezing), the administration of propranolol was perma-
nently discontinued. If these episodes had been observed
within the first 2 days of treatment, these newborns were
included into the control group. The results were analyzed
according to the per-protocol analysis. The study protocol
was approved by the ethics committees of both centers. Writ-
ten informed consent was obtained from the parents.
From February 2010 to May 2012, 56 newborns with ROP
were identified. Three newborns with aggressive posterior
ROP and one newborn with Stage 3 plus were excluded. There-
fore, 52 newborns were enrolled, 26 in the treated and 26 in the
control group (Figure 1; available at www.jpeds.com). One
newborn who initially was enrolled in the treated group
developed serious adverse effects from propranolol on the
first day of treatment and was moved into the control group.
Treated infants were randomized at 67  14 days of postnatal
age (weight 1678  393 g), and the age of newborns in the
control group was 68  17 days (weight 1559  431 g).
Demographic, obstetric, and comorbidities data were not
different between treated and control patients (Table I). A total
of 35 newborns (67.3%) were outborn, and most of them were
transferred for surgery; 15 newborns in the treated group
2 Filippi et al
Vol. -, No. -
(57.7%) and 13 in the control group (50%) had at least one
surgical intervention. Five newborns in the treated group and 2
in the control group had posthemorrhagic hydrocephalus that
was treated with the placement of an Ommaya reservoir
(CSF-Ventricular Reservoirs; Medtronic, Inc, Minneapolis,
Minnesota) and then with a ventriculoperitoneal shunt implant.
The first 4 newborns in the treated group with a GA of 23-
25 weeks and all the newborns in the older GA group were
treated with 0.5 mg/kg/6 hours of propranolol (high dose).
After the appearance of severe adverse events in 3 newborns
in the 23-25 weeks’ group, the dosage was reduced, and the
remaining 8 newborns were treated with 0.25 mg/kg/6 hours
(low dose). Newborns received propranolol for an average of
66.1  31 days (range, 6-90 days): 60.2  35 days (range, 6-90
days) in the 23-25 weeks group, 71.2  28 days (range, 8-90
days) in the older group.
All newborns with GA <32 weeks had ophthalmologic eval-
uations through indirect ophthalmoscopy.18 When ROP in
zone II reached Stage 2 without plus, newborns were enrolled,
and ophthalmologic examinations were scheduled weekly or
more frequently, according to the severity of ROP. The Re-
tCam Imaging System II (Clarity Medical Systems, Pleasan-
ton, California) was used to monitor the progression of the
disease. Propranolol was administered orally as a 2 mg/mL
syrup shortly after feeding. The syrup was prepared by adding
0.2 g of propranolol hydrochloride powder to a 100-mL solu-
tion containing water for injections, sucrose 25 g, anhydrous
citric acid 0.64 g, and sodium citrate tribasic dehydrate 0.2 g.
The syrup is stable in amber glass bottles for at least 1 month at
2-8 C. The treatment was continued until complete develop-
ment of retinal vascularization, although administration was
not permitted for more than 90 days.
The physicians and nurses were aware of the allocated arm,
but the ophthalmologists and data analyst were blinded to the
allocation. When ROP reached Stage 2 or 3 with plus disease,
the newborns were treated with peripheral retinal ablation
performed by conventional laser photocoagulation with
810-nm diode laser (Iridis Quantel Medical, Clermont-
Ferrand, France).17 An intravitreal bevacizumab injection
(Avastin; Roche, Basel, Switzerland) was considered as rescue
treatment for more aggressive ROP with insufficient regres-
sion after laser treatment.19 Laser photocoagulation and/or
intravitreal bevacizumab were performed under general
anesthesia (ketamine 1 mg/kg intravenously or fentanyl 4
mg/kg intravenously). Vitrectomy was considered in new-
borns with ROP Stage 4A to reduce tractional retinal detach-
ment from fibrous proliferation.
Vital signs and respiratory and hemodynamic variables
were monitored continuously (Infinity Delta; Dr€ager Medical
System, Telford, Pennsylvania). Bradycardia was defined as a
single heart-rate decrease below 100 bpm; apnea as a pause
in breathing for more than 20 seconds or less when associated
with desaturation, bradycardia, pallor, or reduced tone; and
hypotension as mean arterial blood pressure less than the
10th percentile for gestation/birth weight and postnatal
age.20 Blood gas analysis, glucose, serum electrolyte levels, liver
and renal function tests, complete blood cell count, and C-
- 2013
reactive protein level were measured once a week for the first 3
weeks after randomization. Standard electrocardiogram and
cardiac ultrasounds were performed weekly for 3 weeks.
Concentrations of propranolol were measured on dried
blood spots by the liquid-chromatography tandem-mass
spectrometry test21 during the first 3 days of treatment and
at the steady state on the 10th day of treatment.22 The pri-
mary indicator of the efficacy of the propranolol treatment
in reducing the progression of ROP was the different inci-
dence of the progression of the disease from Stage 2 ROP
without plus to Stage 2 ROP with plus and to Stage 3 with
or without plus between treated and control groups. Other
indicators included the different incidence of the treatment
with photocoagulation laser, the rescue treatment with beva-
cizumab, the progression to Stage 4 or 5 ROP, and the
different need for vitrectomy. Plasma concentrations of
VEGF and sE-selectin were measured with commercially
available enzyme-linked immunoassay kits (R&D Systems,
Minneapolis, Minnesota) in blood samples collected at
randomization and once weekly for the first 3 weeks after
randomization. All samples were measured in duplicate.
Statistical Analyses
The trial was designed as a pilot study with a limited recruit-
ment. The percentage of any ROP that progresses to more-
severe ROP is around 37%2; we hypothesized that treatment
with propranolol may be able to block the progression of this
disease. To compare the proportion of newborns that pro-
gresses to a more advanced stage, the estimated sample size
for each group, when we considered normal distribution,
an alpha error of 0.05, and a power of 80%, was a minimum
of 22 participants. We used t tests to assess possible differ-
ences in demographic, biochemical, hemodynamic, and res-
piratory variables between the treated and control newborns.
The null hypothesis was accepted with a P > .05. The efficacy
of the treatment was evaluated by means of the risk ratio,23
which is the ratio between the proportion of subjects pro-
gressing to more advanced-stage ROP in the propranolol
group vs the control group. The relative reduction of risk,
which is the reduction percent of events in the treated group
vs the control group event rate, was calculated23 when it
was not possible to calculate the risk ratio. Statistical analyses
were performed with the Statistical Software Program
(McGraw-Hill, New York, New York).24
Results
Safety
In the treated group, one newborn died after 9 days of treat-
ment from intestinal volvulus. In the control group, 2 new-
borns died: one at 61 days after the enrollment from
vancomycin-resistant Enterococcus faecium meningitis and
the other after 87 days from sepsis as the result of infection
with Escherichia coli.
Overall, during the 90 days of the study, the hemodynamic
and respiratory variables did not differ between the treated
and the control groups (Figures 2 and 3; available at www.
Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns, and Perspectives
ORIGINAL ARTICLES
jpeds.com). Heart rate was lower in the treated group for
most of the period; systolic arterial pressure was lower on
days 1, 10, and 21; and diastolic arterial pressure was lower
on day 12. In all cases, the values were within the normal
range. Weight gain was comparable in both groups. Three
weeks after enrollment, 1 newborn in the treated group and
2 newborns in the control group received respiratory
support with nasal continuous positive airway pressure.
During the first 3 weeks of the study, the variables of blood
gas analysis, blood glucose, serum sodium and chloride, liver
and renal function tests, complete blood cell count, and
C-reactive protein levels were comparable in the 2 groups.
Only potassium values were significantly greater in newborns
treated with propranolol after 7, 14, and 21 days of treatment,
although all values were within the normal range (4.2  0.9
mmol/L, 4.0  0.8 mmol/L, and 4.4  1.2 mmol/L in the
treated group vs 3.5  0.5 mmol/L, 3.6  0.5 mmol/L, and
3.6  0.5 mmol/L in the control group, respectively). In addi-
tion, no differences were observed in the laboratory tests as
well as in respiratory and hemodynamic variables between
newborns of GA 23-25 weeks treated with high-dose or
with low-dose propranolol. No newborn required treatment
with inotropes during the study period. No newborns in
either group had electrocardiographic and/or cardiac ultra-
sound abnormalities.
Severe Adverse Effects
A series of apneas, bradycardias, and hypotension were
observed in 3 newborns in the GA 23-25 weeks’ group treated
with high-dose propranolol. Two newborns with GA 24+6
weeks developed these symptoms simultaneously with the
development of an infection (coagulase-negative staphylo-
coccus in one newborn on the first day of treatment and
Enterococcus faecalis in the other on the 20th day of treat-
ment). Another newborn with 24+5 weeks GA on the 36th
day of treatment had these findings, in conjunction with
the induction of general anesthesia before neurosurgery. All
these newborns were resuscitated: one recovered with the
administration of adrenaline, and the others, who were re-
fractory to adrenaline, recovered after receiving terlipressin.
The protocol was amended, and one-half dose was used for
infants in GA 23-25 weeks’ group. However, another
newborn at 23+2 weeks of GA on the 6th day of treatment
had severe apneas and bradycardia after respiratory failure
attributable to massive atelectasis; this newborn responded
to the administration of adrenaline. Severe bradycardia
occurred after 28 days of treatment in one newborn of 30
GA weeks simultaneously with the development of an
H1N1 virus infection. Another 26+4 weeks’ GA infant had se-
vere bronchospasm after 8 days of treatment but responded
to the administration of adrenaline.
Efficacy
There was a significant reduction in the progression to Stage
3 ROP without plus and to Stage 3 plus and, consequently,
there was a reduced need for laser photocoagulation
3
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. -, No. -

Table I. Comparison of baseline characteristics

Group 23-25 weeks Group 26-32 weeks
Treated (n = 12) Control (n = 13) P value Treated (n = 14) Control (n = 13) P value
GA, weeks, mean  SD 24.6  0.8 25.0  0.7 .19 28.2  1.5 27.4  1.5 .15
Birth weight, g, mean  SD 691  127 704  85 .76 860  244 906  208 .60
Male, n (%) 7 (58) 8 (61) .83 10 (71) 7 (58) .36
Cesarean delivery, n (%) 6 (50) 9 (69) .35 13 (93) 12 (92) .96
Stained amniotic fluid, n (%) 0 (0) 1 (8) .37 0 (0) 0 (0) 1.00
Outborn, n (%) 9 (75) 9 (69) .76 8 (57) 9 (69) .53
Apgar score, 1 min, mean  SD 3.1  2.3 4.7  1.8 .07 5.2  2.3 5.1  2.3 .85
Apgar score, 5 min, mean  SD 6.4  1.5 7.5  0.9 .05 7.5  1.5 7.4  1.0 .82
White race, n (%) 12 (100) 11 (85) .17 13 (93) 12 (92) .96
Age at enrollment, days, mean  SD 74.2  12.2 75.5  16.9 .83 61.5  13.3 61.5  13.9 .99
Postmenstrual age at the enrollment, weeks, mean  SD 35.2  1.8 35.8  2.3 .48 36.4  2.1 36.2  1.7 .77
Weight at enrollment, g, mean  SD 1566  314 1539  351 .84 1676  451 1604  480 .69
Respiratory support at the enrollment, n (%) 7 (58) 6 (46) .56 3 (21) 4 (31) .60
Survival, n (%) 11 (92) 11 (85) .61 14 (100) 13 (100) 1.00
Comorbidities and cointerventions
Respiratory distress syndrome, n (%) 12 (100) 13 (100) 1.00 13 (93) 13 (100) .34
Surfactant treatment, n (%) 12 (100) 13 (100) 1.00 10 (71) 10 (77) .76
Oxygen exposure, n (%) 12 (100) 12 (92) 1.00 11 (79) 10 (77) .92
Duration of oxygen exposure (days), median (range) 80.5 (31-181) 58.5 (0-126) .25 20.5 (0-102) 32.6 (0-167) .38
Bronchopulmonary dysplasia, n (%)* 9 (75) 5 (38) .07 5 (36) 4 (31) .79
Candida sepsis, n (%) 3 (25) 3 (23) .91 1 (7) 1 (8) .96
Other sepsis, n (%) 8 (67) 7 (54) .53 4 (29) 4 (31) .90
Newborns transfused with red blood cells, n (%) 12 (100) 13 (100) 1.00 11 (79) 11 (85) .70
Number of red blood cell transfusions, median (range) 7.5 (4-9) 7 (3-9) .27 1.5 (0-10) 4 (0-11) .36
Intraventricular hemorrhage, grade 3-4, n (%) 6 (50) 6 (46) .85 2 (14) 2 (15) .94
Posthemorrhagic hydrocephalus, n (%) 2 (17) 2 (15) .93 3 (21) 0 (0) .08
Neurosurgery, n (%) 2 (17) 2 (15) .93 3 (21) 0 (0) .08
Surgical closure of patent ductus arteriosus, n (%) 5 (42) 4 (31) .59 2 (14) 2 (15) .94
Surgery for intestinal disease, n (%) 5 (42) 5 (38) .88 6 (43) 7 (54) .58
*Oxygen requirement at 36 weeks’ postmenstrual age.

treatment in newborns receiving propranolol (Table II). In
the treated group, the progression to Stage 3 plus was
observed in an infant with posthemorrhagic hydrocephalus
and in another who had stopped receiving propranolol
after 6 days of treatment because of apnea and bradycardia.
The number of newborns and eyes treated with
bevacizumab was lower in the treated group than in the
control group, and no newborns progressed to Stage 4. In
the majority of newborns after 10-15 days of treatment, the
ridge tended to turn pale and a second, more peripheral
ridge appeared usually on 2 or 3 quadrants. After 1 month,
the peripheral ridge was more evident than the first one,
which disappeared after 2 months of therapy, leaving just a
thin demarcation line (Figure 4, A; www.jpeds.com). After
approximately 2 months, the normal retinal vessels also
passed over the peripheral ridge, which was reduced in
width and height (Figure 4, B).
Plasma Propranolol, VEGF, and sE-Selectin Levels
Plasma propranolol concentrations during the first 3 days
and those at the 10th day are shown in Figure 5 (available
at www.jpeds.com). Plasma VEGF concentrations were
highly variable in both groups but not significantly
different between treated and control newborns and
between those treated with high doses and low doses, at
any time (Table III; available at www.jpeds.com). Plasma
sE-selectin concentrations were significantly lower in
treated newborns starting from 7 days of treatment
(Figure 6). The comparison between sE-selectin con-
centrations at randomization and after 21 days of
treatment showed a significant reduction in the treated
neonates (from 49.7  7.1 ng/mL to 39.8  10.3 ng/mL;
P = .01) but not in the control group (from 56.7  7.7 to
60.3  14.5 ng/mL; P = .57). No differences were observed
between newborns treated with low doses and high doses.
Discussion
This clinical trial was stimulated by studies in animal models
that revealed the role of the adrenergic system in the develop-
ment of proliferative retinopathy. We recently demonstrated
in C57Bl/6J mice with OIR that treatment with propranolol
down-regulates retinal levels of angiogenic factors such as
VEGF and reduces retinal neovascularization induced by hyp-
oxia.12 The mechanism occurs through the block of b2-AR,
predominantly localized in M} uller cells.13 A similar effect
was observed with the b2-AR desensitization after the pro-
longed administration of isoproterenol.25 In this mice strain,
retinal levels of b1- and b2-ARs were not modified by hypoxia,
whereas b3-ARs, localized in engorged retinal tufts,
were increased by 120%.13 The antiangiogenic efficacy of
propranolol was confirmed in C57Bl/6J mice with choroidal
neovascularization.26 However, in129S6 mice, a strain with
greater susceptibility to retinal neovascularization induced

4 Filippi et al
- 2013 ORIGINAL ARTICLES

Table II. Ophthalmologic outcome

Treated (n = 25) Control (n = 26) P value Risk ratio (95% CI) Relative reduction of risk, %
Progression to Stage 2 ROP with plus, n (%)
Newborns 0/25 (0) 1/26 (4) .51 nd 100
Eyes 0/50 (0) 2/52 (4) .31 nd 100
Progression to Stage 3 ROP without plus, n (%)
Newborns 9/25 (36) 18/26 (69) .03 0.520 (0.47-0.58) 48
Eyes 15/50 (30) 31/52 (60) .01 0.503 (0.47-0.54) 50
Progression to Stage 3 ROP with plus, n (%)
Newborns 4/25 (16) 10/26 (38) .09 0.416 (0.31-0.58) 58
Eyes 8/50 (16) 19/52 (37) .03 0.438 (0.38-0.50) 56
Treatment with laser photocoagulation, n (%)
Newborns 4/25 (16) 10/26 (38) .09 0.416 (0.31-0.58) 58
Eyes 8/50 (16) 19/52 (37) .03 0.438 (0.38-0.50) 56
Rescue treatment with bevacizumab, n (%)
Newborns 2/25 (8) 5/26 (19) .26 0.480 (0.29-0.79) 52
Eyes 3/50 (6) 7/52 (13) .22 0.444 (0.31-0.64) 55
Progression to Stage 4 ROP, n (%)
Newborns 0/25 (0) 4/26 (15) .14 nd 100
Eyes 0/50 (0) 4/52 (8) .14 nd 100
Vitrectomy, n (%)
Newborns 0/25 (0) 0/26 (0) .98 nd nd
Eyes 0/50 (0) 0/52 (0) .98 nd nd

nd, not determined.

by hypoxia,27 hypoxia up-regulated b3-ARs>10-fold, and angiomas, the most common vascular tumor of infancy,8
propranolol, which is only minimally active in blocking by inhibition of VEGF expression.36 Although propranolol
b3-ARs,28 failed to reduce the VEGF levels and neovasculari- generally is well tolerated in newborns and infants, premature
zation.29 These results suggest that b3-ARs also play a patho- infants have a high risk of associated complications (sepsis,
genic role in retinal angiogenesis that could be related to apneas, respiratory bronchodysplasia) that could make this
different genetic backgrounds of mouse strains.30 In mouse drug unsafe. For this reason, the main objective of this study
retinal explants, we demonstrated recently that b3-ARs can was to evaluate its safety.37
modulate VEGF release in response to hypoxia via the nitric This trial demonstrates that, in stable newborns, treat-
oxide pathway, suggesting that blocking this signaling ment with propranolol is definitely well tolerated, as
pathway may be effective in reducing VEGF-stimulated demonstrated by the stability of the biochemical, hemato-
retinal angiogenesis.14 logic, hemodynamic, and respiratory variables. The plasma
The concern that propranolol could damage the develop- VEGF levels were not modified by the administration of
ment of the premature infant’s brain as the result of its effects propranolol, regardless of the dose. These results are
on VEGF is justified by experimental evidence showing that, in consistent with previous data reporting unchanged VEGF
mice, VEGF participates in brain morphogenesis31 and severe plasma levels in infants with ROP and suggest that
reductions in VEGF levels lead to degeneration of the cerebral
cortex.32 In experiments in animals, propranolol only reduced
VEGF overproduction in the hypoxic retina of mice with OIR
without affecting the VEGF levels in the normoxic retina of
control mice, suggesting different mechanisms of VEGF regu-
lation in normoxic and hypoxic conditions.12 In addition, b-
AR blockade did not influence VEGF levels in the brain, lungs,
or heart, where the OIR model did not induce hypoxia or
VEGF up-regulation.12 However, subcutaneous propranolol
at 20 mg/kg in rats blocks the acquisition of a conditioned
odor preference and associated neural responses.33 In chicks,
intracerebral propranolol injections cause memory loss.34
Although these effects were observed with doses 10-20 times
greater than those used in our study, propranolol had dose-
dependent effects on odor-preference learning.35 Therefore,
it may be important to monitor plasma VEGF levels and neu-
rodevelopmental in all infants treated with b-blockers.
Figure 6. Mean plasma concentrations of sE-selectin during
Propranolol has been used for several decades in young in- the first 3 weeks of the study in treated group and in control
fants for a variety of cardiovascular indications,22 but it was group.
serendipitously shown to induce involution of infantile hem-
Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns, and Perspectives 5
THE JOURNAL OF PEDIATRICS  www.jpeds.com
pathogenic retinal angiogenesis is mostly driven by local
VEGF synthesis.38
To rule out possible adverse effects of propranolol admin-
istration on brain development, the neurodevelopmental
outcome as well as the visual function (ocular motricity, vi-
sual acuity, visual field, stereopsis) of newborns enrolled in
this trial was evaluated by means of neurologic and behav-
ioral standardized tests.37 Preliminary data at 12 months of
corrected age show that all infants treated with propranolol
have good development of visual function except for stra-
bismus that persists in 25% of cases. Approximately 80% of
the treated patients have visual acuity within the normal
range, 60% of the control group have normal visual acuity,
and some were totally blind. Regarding the ocular motricity,
infants treated with propranolol are better able to keep a
smooth fixation on a slowly moving target for a wider angle
than patients in the control group. Nystagmus also persists in
approximately 30% of the control group and in 10% of
treated infants.
When propranolol was administered to unstable new-
borns (during sepsis, respiratory impairment, or after anes-
thesia induction), there was a high risk of hypotension and
bradycardia that are sometimes unresponsive to catechol-
amines. However, the severe hypotension was reversed
with terlipressin, an agent active on different receptors.39
In contrast to infants and children, in whom adverse events
after treatment with propranolol are usually mild and not
life-threatening, preterm newborns appear to have a greater
risk of developing severe adverse effects when complications
frequently associated with prematurity, like sepsis, develop
during treatment with b-blockers. Because of the increased
stiffness of their myocardium, neonates possess a limited
ability to increase their cardiac output by increasing stroke
volume, and their cardiac output is significantly dependent
on heart rate.40 Because myocardial depression is a well-
known manifestation of organ dysfunction in sepsis, treat-
ment with b-blockers may prevent the only compensatory
mechanism to increase the myocardial function.
Our study population showed a high incidence of comor-
bidities that usually promote ROP progression
(Table I)41,42 and a high percentage of infants in the control
group progressed to more advanced stages of ROP. Sixty-
nine percent of control newborns developed Stage 3, and
38.5% progressed to Stage 3 plus and were treated with laser
photocoagulation. Another possible explanation for the
high incidence of severe ROP in our population was the use
of a high target range for oxygen saturation. Recent trials
have demonstrated that despite reducing the risk of severe
retinopathy among survivors, a low target range of oxygen
saturation (85%-89%) increased the in-hospital mortality
among extremely premature infants.43 However, the risk of
ROP progression was significantly lower in the group
treated with propranolol, even though newborns in the 23-
25 weeks’ group received oxygen for longer time, and
newborns in the 26-32 weeks’ group had a greater incidence
of posthemorrhagic hydrocephalus requiring neurosurgery.
The number of newborns receiving rescue treatment with
6 Filippi et al
Vol. -, No. -
bevacizumab and progressing to Stage 4 was lower in the
treated group, but the total number of patients was too low
for statistical significance. A similar result recently was
reported in adult patients with age-related macular
degeneration cotreated with b-blockers.44 Another sign of
the antiangiogenic effect of propranolol was the reduced
level of sE-selectin, an inducible endothelial leukocyte
adhesion molecule expressed on the surface of endothelial
cells, in treated newborns and it is directly involved in
angiogenesis and capillary morphogenesis, the plasma levels
of sE-selectin are closely related to ROP.45
This trial has some limitations. First, it was planned as a pi-
lot study with a small number of enrolled patients. Second,
the study was not placebo-controlled, and only the ophthal-
mologists were blinded. This strategy was chosen to identify
any possible adverse effects and immediately discontinue
the administration of propranolol. Many uncertainties still
remain. First, propranolol was chosen in doses that usually
do not produce adverse effects but are effective in the treat-
ment of hemangiomas.8 However, it is unknown at what con-
centration, when administered orally, propranolol reaches
the retina. Second, it was not possible to estimate the optimal
start and duration of administration.
In conclusion, our data suggest that treatment with pro-
pranolol is effective in counteracting the progression of
ROP and add new perspectives for the treatment of this
and other proliferative retinopathies. Nevertheless, the
high incidence of adverse effects in newborns suggests
that its systemic administration is not sufficiently safe. A
future objective will be to explore and develop topical de-
livery systems to ensure high retinal and low plasma con-
centrations of propranolol. In this regard, we recently
described how application of propranolol as eye drops in
mice promotes the recovery of OIR,15 suggesting that the
topical application of propranolol could represent an alter-
native delivery route to systemic administration. n
We are grateful to the nursing staff of the Neonatal Intensive Care
Units of A. Meyer University Children’s Hospital, Florence, and of
Fondazione IRCCS C
a Granda Ospedale Maggiore Policlinico, Milan,
Italy for their assistance in conducting this study, and to Dr Giovanni
Casini for his expert editorial assistance.
Submitted for publication Jan 17, 2013; last revision received Jul 15, 2013;
accepted Jul 31, 2013.
Reprint requests: Filippi Luca, MD, Neonatal Intensive Care Unit, Medical
Surgical Fetal-Neonatal Department, “A. Meyer” University Children’s
Hospital, Viale Pieraccini, 24 I-50139 Florence, Italy. E-mail: l.filippi@meyer.it
References
1. Gilbert C. Retinopathy of prematurity: a global perspective of the epi-
demics, population of babies at risk and implications for control. Early
Hum Dev 2008;84:77-82.
2. Good WV, Hardy RJ, Dobson V, Palmer EA, Phelps DL, Quintos M,
et al. The incidence and course of retinopathy of prematurity: findings
from the early treatment for retinopathy of prematurity study. Pediatrics
2005;116:15-23.
3. Penn JS, Madan A, Caldwell RB, Bartoli M, Caldwell RW, Hartnett ME.
Vascular endothelial growth factor in eye disease. Prog Retin Eye Res
2008;27:331-71.
- 2013
4. Good WV, Hardy RJ, Wallace DK, Bremer D, Rogers DL,
Siatkowski RM, et al. b-Blocking and racial variation in the severity of
retinopathy of prematurity. Arch Ophthalmol 2012;130:117-8.
5. Saunders RA, Donahue ML, Christmann LM, Pakalnis AV, Tung B,
Hardy RJ, et al. Racial variation in retinopathy of prematurity. The Cryo-
therapy for Retinopathy of Prematurity Cooperative Group. Arch Oph-
thalmol 1997;115:604-8.
6. Liggett SB, Cresci S, Kelly RJ, Syed FM, Matkovich SJ, Hahn HS, et al. A
GRK5 polymorphism that inhibits beta-adrenergic receptor signaling is
protective in heart failure. Nat Med 2008;14:510-7.
7. Baker JG. The selectivity of beta-adrenoceptor antagonists at the human
beta1, beta2 and beta3 adrenoceptors. Br J Pharmacol 2005;144:317-22.
8. Leaute-Labr
eze C, Dumas de la Roque E, Hubiche T, Boralevi F,
Thambo JB, Ta€ıeb A. Propranolol for severe hemangiomas of infancy.
N Engl J Med 2008;358:2649-51.
9. Praveen V, Vidavalur R, Rosenkrantz TS, Hussain N. Infantile hemangi-
omas and retinopathy of prematurity: possible association. Pediatrics
2009;123:e484-9.
10. Smith LE, Wesolowski E, McLellan A, Kostyk SK, D’Amato R,
Sullivan R, et al. Oxygen-induced retinopathy in the mouse. Invest Oph-
thalmol Vis Sci 1994;35:101-11.
11. Chen J, Smith LE. Retinopathy of prematurity. Angiogenesis 2007;10:
133-40.
12. Ristori C, Filippi L, Dal Monte M, Martini D, Cammalleri M,
Fortunato P, et al. Role of the adrenergic system in a mouse model
of oxygen-induced retinopathy: antiangiogenic effects of beta-
adrenoreceptor blockade. Invest Ophthalmol Vis Sci 2011;52:155-70.
13. Martini D, Monte MD, Ristori C, Cupisti E, Mei S, Fiorini P, et al. Anti-
angiogenic effects of b2 -adrenergic receptor blockade in a mouse model
of oxygen-induced retinopathy. J Neurochem 2011;119:1317-29.
14. Dal Monte M, Filippi L, Bagnoli P. Beta3-adrenergic receptors modulate
vascular endothelial growth factor release in response to hypoxia
through the nitric oxide pathway in mouse retinal explants. Naunyn
Schmiedebergs Arch Pharmacol 2013;386:269-78.
15. Dal Monte M, Casini G, la Marca G, Isacchi B, Filippi L, Bagnoli P. Eye
drop propranolol administration promotes the recovery of oxygen-
induced retinopathy in mice. Exp Eye Res 2013;111C:27-35.
16. International Committee for the Classification of Retinopathy of Prema-
turity. The International Classification of Retinopathy of Prematurity re-
visited. Arch Ophthalmol 2005;123:991-9.
17. Early Treatment For Retinopathy Of Prematurity Cooperative Group.
Revised indications for the treatment of retinopathy of prematurity: re-
sults of the early treatment for retinopathy of prematurity randomized
trial. Arch Ophthalmol 2003;121:1684-94.
18. Section on Ophthalmology American Academy of Pediatrics, American
Academy of Ophthalmology, American Association for Pediatric
Ophthalmology and Strabismus. Screening examination of premature
infants for retinopathy of prematurity. Pediatrics 2006;117:572-6.
19. Jalali S, Balakrishnan D, Zeynalova Z, Padhi TR, Rani PK. Serious
adverse events and visual outcomes of rescue therapy using adjunct bev-
acizumab to laser and surgery for retinopathy of prematurity. The Indian
Twin Cities Retinopathy of Prematurity Screening database Report
number 5. Arch Dis Child Fetal Neonatal Ed 2013;98:F327-33.
20. Nuntnarumit P, Yang W, Bada-Ellzey HS. Blood pressure measurements
in the newborn. Clin Perinatol 1999;26:981-96.
21. Della Bona ML, Malvagia S, Villanelli F, Giocaliere E, Ombrone D,
Funghini S, et al. A rapid liquid chromatography tandem mass
spectrometry-based method for measuring propranolol on dried blood
spots. J Pharm Biomed Anal 2013;78-79:34-8.
22. Filippi L, Cavallaro G, Fiorini P, Malvagia S, Della Bona ML,
Giocaliere E, et al. Propranolol concentrations after oral administration
in term and preterm neonates. J Matern Fetal Neonatal Med 2013;26:
833-40.
23. Glantz SA. Primer of bostatistics. 6th ed. New York: McGraw-Hill; 2005.
24. Glantz SA. Statistical software program. Version 6.0. New York:
McGraw-Hill; 2005.
Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns, and Perspectives
ORIGINAL ARTICLES
25. Dal Monte M, Martini D, Latina V, Pavan B, Filippi L, Bagnoli P. Beta-
adrenoreceptor (b-AR) agonism influences retinal responses to hypoxia
in a mouse model of retinopathy of prematurity. Invest Ophthalmol Vis
Sci 2012;53:2181-92.
26. Lavine JA, Sang Y, Wang S, Ip MS, Sheibani N. Attenuation of choroidal
neovascularization by b2-adrenoreceptor antagonism. JAMA Ophthal-
mol 2013;131:376-82.
27. Chan CK, Pham LN, Zhou J, Spee C, Ryan SJ, Hinton DR. Differential
expression of pro- and antiangiogenic factors in mouse strain-
dependent hypoxia-induced retinal neovascularization. Lab Invest
2005;85:721-33.
28. Hoffmann C, Leitz MR, Oberdorf-Maass S, Lohse MJ, Klotz KN.
Comparative pharmacology of human beta-adrenergic receptor sub-
types—characterization of stably transfected receptors in CHO cells.
Naunyn Schmiedebergs Arch Pharmacol 2004;369:151-9.
29. Chen J, Joyal JS, Hatton CJ, Juan AM, Pei DT, Hurst CG, et al. Propran-
olol inhibition of b-adrenergic receptor does not suppress pathologic
neovascularization in oxygen-induced retinopathy. Invest Ophthalmol
Vis Sci 2012;53:2968-77.
30. Filippi L, Monte MD, Bagnoli P. Different efficacy of propranolol in
mice with oxygen-induced retinopathy: could differential effects of pro-
pranolol be related to differences in mouse strains? Invest Ophthalmol
Vis Sci 2012;53:7421-3.
31. Raab S, Beck H, Gaumann A, Y€ uce A, Gerber HP, Plate K, et al. Impaired
brain angiogenesis and neuronal apoptosis induced by conditional ho-
mozygous inactivation of vascular endothelial growth factor. Thromb
Haemost 2004;91:595-605.
32. Haigh JJ, Morelli PI, Gerhardt H, Haigh K, Tsien J, Damert A, et al.
Cortical and retinal defects caused by dosage-dependent reductions in
VEGF-A paracrine signaling. Dev Biol 2003;262:225-41.
33. Sullivan RM, Stenwalt G, Nasr F, Lemon C, Wilson DA. Association of
an odor with activation of olfactory bulb noradrenergic beta-receptors
or locus coeruleus stimulation is sufficient to produce learned approach
responses to that odor in neonatal rats. Behav Neurosci 2000;114:957-62.
34. Gibbs ME, Hutchinson DS, Summers RJ. Noradrenaline release in the
locus coeruleus modulates memory formation and consolidation; roles
for a- and b-adrenergic receptors. Neuroscience 2010;170:1209-22.
35. Sullivan RM, Wilson DA, Leon M. Norepinephrine and learning-
induced plasticity in infant rat olfactory system. J Neurosci 1989;9:
3998-4006.
36. Chim H, Armijo BS, Miller E, Gliniak C, Serret MA, Gosain AK. Pro-
pranolol induces regression of hemangioma cells through HIF-
1a mediated inhibition of VEGF-A. Ann Surg 2012;256:146-56.
37. Filippi L, Cavallaro G, Fiorini P, Daniotti M, Benedetti V, Cristofori G,
et al. Study protocol: safety and efficacy of propranolol in newborns with
Retinopathy of Prematurity (PROP-ROP): ISRCTN18523491. BMC Pe-
diatr 2010;10:83.
38. Pieh C, Agostini H, Buschbeck C, Kr€ uger M, Schulte-M€ onting J,
Zirrgiebel U, et al. VEGF-A, VEGFR-1, VEGFR-2 and Tie2 levels in
plasma of premature infants: relationship to retinopathy of prematurity.
Br J Ophthalmol 2008;92:689-93.
39. Filippi L, Gozzini E, Daniotti M, Pagliai F, Catarzi S, Fiorini P. Rescue
treatment with terlipressin in different scenarios of refractory hypoten-
sion in newborns and infants. Pediatr Crit Care Med 2011;12:e237-41.
40. Costello JM, Franklin WH. Preoperative and postoperative care of
the infant with congenital heart disease. In: MacDonald MG,
Seshia MMK, Mullett MD, eds. Avery’s neonatology pathophysiology
& management of the newborn. 6th ed. Philadelphia: Lippincott
Williams & Wilkins; 2005. p. 738.
41. Bourla DH, Gonzales CR, Valijan S, Yu F, Mango CW, Schwartz SD.
Association of systemic risk factors with the progression of laser-
treated retinopathy of prematurity to retinal detachment. Retina 2008;
28:S58-64.
42. Sood BG, Madan A, Saha S, Schendel D, Thorsen P, Skogstrand K, et al.
Perinatal systemic inflammatory response syndrome and retinopathy of
prematurity. Pediatr Res 2010;67:394-400.
7
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. -, No. -

43. SUPPORT Study Group of the Eunice Kennedy Shriver NICHD
Neonatal Research Network, Carlo WA, Finer NN, Walsh MC,
Rich W, Gantz MG, Laptook AR, et al. Target ranges of oxygen satura-
tion in extremely preterm infants. N Engl J Med 2010;362:1959-69.
44. Montero JA, Ruiz-Moreno JM, Sanchis-Merino E, Perez-Martin S. Sys-
temic beta-blockers may reduce the need for repeated intravitreal injec-
tions in patients with age-related macular degeneration treated by
bevacizumab. Retina 2013;33:508-12.
45. Pieh C, Kr€uger M, Lagr
eze WA, Gimpel C, Buschbeck C, Zirrgiebel U,
et al. Plasma sE-selectin in premature infants: a possible surrogate
marker of retinopathy of prematurity. Invest Ophthalmol Vis Sci 2010;
51:3709-13.

8 Filippi et al
- 2013 ORIGINAL ARTICLES

Figure 1. Enrollment, randomization, and analysis of the 52 study newborns.

Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns, and Perspectives 8.e1
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. -, No. -

Figure 2. Hemodynamic variables during the 90 days of the study.

8.e2 Filippi et al
- 2013 ORIGINAL ARTICLES

Figure 3. Respiratory variables during the 90 days of the study. FiO2, fraction of inspired oxygen; SaO2, oxygen saturation.

Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns, and Perspectives 8.e3
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. -, No. -

Figure 4. A, Appearance, after approximately 1 month of

treatment, of a second peripheral ridge that gradually be-
comes more evident than the first one. B, After about 2
months, the normal retinal vessels pass over the peripheral
ridge.

8.e4 Filippi et al
- 2013 ORIGINAL ARTICLES

Figure 5. Plasma propranolol concentrations measured by dried blood spots in the 14 newborns with GA 26-32 weeks, and in
the 4 newborns with GA 23-25 weeks, treated with 0.5 mg/kg/6 hours of propranolol (high dose) and in the 8 newborns with GA
23-25 weeks treated with 0.25 mg/kg/6 hours of propranolol (low dose) A, during the first 3 days and B, in the 10th day of
treatment.

Oral Propranolol for Retinopathy of Prematurity: Risks, Safety Concerns, and Perspectives 8.e5
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. -, No. -

Table III. Median plasma concentrations of VEGF during the first 3 weeks of the study*
Treated (n = 25)
Serum VEGF levels, pg/mL Control (n = 26) All newborns (n = 25) High dose (n = 18) Low dose (n = 7) P value
T0, median (range) 197 (79-1256) 127 (34-335) .14
124 (34-335) 107 (88-258) .98
T7, median (range) 208 (140-1347) 126 (64-863) .36
185 (93-863) 116 (64-165) .30
T14, median (range) 123 (14-668) 280 (38-2394) .21
280 (69-751) 264 (38-2394) .91
T21, median (range) 150 (59-1210) 155 (52-1601) .92
155 (82-365) 314 (52-1601) .16

T0, time of randomization; T7, T14, T21, after 7, 14, 21 days of treatment, respectively.
*Data of treated newborns are presented also distinguishing newborns treated with a high or low dose of propranolol.

8.e6 Filippi et al