Oral Propranolol Versus Placebo for Early Stages of Retinopathy of Prematurity: A Randomized and Prospective Study

Retinopathy of prematurity (ROP) affects the retinal microvasculature, mostly of very-low-birth-weight (VLBW: < 1500g) premature infants, and is a significant complication of extreme prematurity leading sometimes to devastating consequences. Although ROP is usually mild with no harm, it happens not very rarely to be aggressive causing neo-vascularization (NV) in the immature retina that at times can progress to severe fibrovascular proliferation, retinal detachment and blindness (1). Major risk factors for ROP are low gestational age, low birth weight, hyperoxia, respiratory distress syndrome (RDS) and intraventricular hemorrhage (IVH) (2) as well as postnatal steroid therapy (3). Of note is one report showing that the use of beta-blocking agents by the mother before birth was found to be associated with the development of ROP (4). However, so far no one reported a similar effect of the postnatal use of beta blockers on ROP.

The incidence of ROP is inversely related to gestational age (GA) and birth weight (BW). The condition develops in 51% of infants with a birth weight (BW) <1700 g (5). In infants weighing less than 1250 g, 50% show some evidence of ROP and 10% progress to stage III ROP. According to the Israeli VLBW-Database in 2007, 23.9% of infants develop ROP (all stages), while 4.8% develop severe ROP (stage III-IV) (4). Worldwide, at least 50,000 children are blind from ROP (2,7). In South Africa it accounts for 10.6% of cases of childhood blindness (8). In the US, annually, 500-700 children become blind due to ROP, and 2100 infants will be affected by cicatricial sequelae, such as myopia, strabismus, as well as late-onset retinal detachment (1).

LASER photocoagulation of the ischemic retina is the therapy of choice for moderate to severe ROP and is required in 19.8%, 7.7%. 1.5% and 0.6% of infants weighing 500-749g, 750-999g, 1000-1249g, 1250-1499g, respectively (2). In Israel, 4% of VLBW infants needed LASER photocoagulation or cryo therapy during 2007 (6).

The pathogenesis of ROP is multifactorial and two pathogenetic theories have been proposed:

(A) One-phase theory: Mesenchymal spindle cells when exposed to extrauterine hyperoxia, develop gap junctions that interfere with normal vascular formation and trigger a neovascular response (9).

(B) Two-phase theory: The first phase (hyperoxic phase, vaso-obliterative), consists of retinal vasoconstriction and irreversible capillary endothelial cell damage. As the retinal area that is supplied by the affected vessels becomes ischemic, angiogenic factors, such as vascular endothelial growth factor (VEGF) are produced by mesenchymal spindle cells in that ischemic retina to provide neo vascularization (NV) channels (second phase, vaso-proliferative)(10).

Increasing evidence supports the key role of VEGF in the pathogenesis of ROP, wherein VEGF is down-regulated in the vaso-obliterative first phase and up-regulated in the vaso-proliferative second phase (11). Numerous studies have been performed in an animal model of oxygen-induced retinopathy (OIR), wherein newborn rats, mice, kitten and beagle puppies were exposed to 75-100% O2 for 5 days starting at day 7 of life (11). ROP usually develops in 100% of the O2-exposed rats (12).

The expression of various angiogenetic and inflammation genes has been studied in oxygen-induced retinopathy (OIR). Sato et al (13) investigated the expression of 94 genes in OIR using microarray analysis and reverse transcriptase-polymerase chain reaction (RT-PCR). They observed that: (a) Inflammation genes were up-regulated at days 12-13 of life when the degree of both central avascular area and central vasoconstriction were maximal; this up-regulation continued until day 21 of life; (b) Extra retinal vascularization was most noticeable at days 16-17 of life, when angiogenesis genes (VEGF-A, angiopoietin-2) were at their highest expression.

There is also increasing evidence of up-regulation of VEGF by sympathomimetic agents. In this regard, norepinephrine has been shown to stimulate myocardial angiogenesis in rats (14). In cultured retinal endothelial cells, Steinle et al (15) showed that significantly increased expression of beta-3 receptors could promote migration and proliferation (two markers of angiogenic phenotype) of retinal endothelial cells.

In addition, in cancer cell cultures, catecholamines (norepinephrine and epinephrine) induced an increase of VEGF expression in a tissue culture of nasopharyngeal carcinoma, an effect that was blocked by propranolol (prop) (a non-selective beta blocker) (16). Evidence exists for norepinephrine-induced invasiveness with increased VEGF in human pancreatic cell lines, could also be blocked by prop (17). Blockade of these effects by prop raised the prospects of a possible chemo-prevention of vascularization-rich tumors by propranolol.

Recent studies have shown that administration of beta-blockers (both locally and systemic) can mitigate NV, probably by down regulation of VEGF. In an animal model of OIR, topical timolol (a beta blocker) prevented the development of OIR in 40% of rats and mitigated the severity of OIR in the remaining 60% of rats that had developed OIR (12,18). In addition, timolol had a protective effect, whereby NV occurred in 65% of timolol-treated as compared to 100% NV in untreated rats (19). Furthermore, VEGF expression was lower in timolol-treated rats than in controls (room air). In contrast, Zheng et al (20) found no effect of prop on the VEGF protein and on messenger ribonucleic acid (mRNA) expression in the retina of diabetic rats with retinopathy.

ROP and infantile hemangiomas (a rather common phenomena in premature infants) supposedly share the same pathogenetic role of angiogenic factors such as VEGF (21). In a recent report by Praveen et al (22), a possible association between ROP and infantile hemangiomas at discharge was studied in premature infants weighing <1250 g. Infantile hemangiomas were found to be independently associated with any stage of ROP: infantile hemangiomas were present in 16.8% of neonates with ROP as compared with 6.7% of those without ROP. However, neither the size nor the number of infantile hemangiomas showed any association with the severity of ROP.

The above-mentioned published findings point to a VEGF-mediated pathogenesis of both ROP and infantile hemangiomas, wherein VEGF expression is reportedly up-regulated by sympathomimetic agents and blocked by beta blockers. Furthermore, on the clinical scene, the usefulness of prop in mitigating the progression of infantile hemangiomas has been recently reported (23-32). Infants with severe or life-threatening hemangiomas were successfully treated with prop, with no adverse effects. One potential explanation for the effect of prop on hemangiomas includes: (a) vasoconstriction, or (b) decreased expression of VEGF and basic fibroblast growth factor (bFGF) genes through the down-regulation of RAF-mitogen-activated protein kinase pathway (33) (which explains the progressive improvement of hemangioma), or (c) a triggering of capillary endothelial cells apoptosis (34).

Prop administration has been observed to be safe in infants and toddlers (23-32, 35). Love et al (35) found that after 40 years of clinical use in infants and toddlers, there is no documented case of death or serious cardiovascular disease as a direct result of exposure to beta-blockers. Furthermore, prop was also reported to be safe when given to premature infants for treatment of neonatal thyrotoxicosis, neonatal arrhythmia or life-threatening hemangioma (32, 36-39). In five extreme-low-birth weight infants (weight <1000 g), the use of prop for neonatal thyrotoxicosis was beneficial and had no adverse effects (36). The safety of prop use was also reported in a 34-week infant (37) and a for 37-week infant (38) with thyrotoxicosis, and also in a 35-week infant with neonatal arrhythmia (39). Furthermore, a 28-week premature infant was treated for18 weeks with prop for a thoracic hemangioma without untoward effects (32).