Retinopathy of prematurity (ROP) is a potentially blinding disease caused by pathologic angiogenesis that occurs in the incompletely vascularized retina of preterm newborns. Despite current therapeutic strategies, ROP still represents a leading cause of potentially avoidable visual impairment and blindness in childhood. More than 30,000 preterm infants become blind or visually impaired from ROP each year worldwide [
1]. In the 1940s, the so-called “first ROP epidemic” was related to the widespread use of unrestricted oxygen supplementation; the second “ROP epidemic” occurred in high-income countries in the 1970s and it was related to the increasing survival rate at lower gestational age (GA) [
2‐
4]. In the early 1990s, an emerging epidemic of blindness due to ROP was also recorded in middle-income countries [
5]. Currently, Asia is the region presenting the highest incidence of blindness due to ROP, followed by Latin America, where some countries account for an incidence of blindness/severe visual impairment related to ROP that is 2.4 times higher than in highly industrialized countries [
1,
6,
7]. Therefore, the detection of a new inexpensive and easily affordable treatment strategy may be a relevant issue of global interest. Prematurity and low birth weight are the main factors associated with ROP, although other factors (i.e. respiratory failure, fetal hemorrhage, intra-ventricular hemorrhage, blood transfusions, hyperglycemia, sepsis, necrotizing enterocolitis) have been described as contributing factors to ROP development [
8,
9].
Physiologically, retinal blood vessels development begins at the optic disc during the fourth month of gestation in the hypoxic uterine environment and is completed at approximately 40 weeks of gestational age. The pathogenesis of ROP has not yet been totally clarified, but the most validated hypothesis describes two different postnatal phases [
10]. During the first phase, the loss of the placenta and the exposure to extrauterine relative hyperoxia are associated with low levels of Vascular Endothelial Growth Factor (VEGF) and Insulin-like Growth Factor 1 (IGF-1), resulting in a cessation of retinal vascularization [
11‐
14]. In fact, oxygen induces retinal vasoconstriction, prevents retinal vessel growth and therefore still represents one of the main determinant of ROP development [
15]. During the second phase, the retinal maturation and the development of relative hypoxia stimulate the VEGF and IGF-1 expression, causing a shift to a proliferative phase, which is characterized by an abnormal angiogenesis [
16‐
18].
For a long time an oxygen saturation level lower than 90% has been suggested to reduce ROP risk. However, the recent demonstration that a higher oxygen saturation (91–95%) correlates with an improved survival represents an actual dilemma because, unfortunately, it induces a higher risk of ROP development [
15]. Apart from oxygen tension, which is the main factor promoting the expression of angiogenic growth factors in proliferative retinopathies, other mechanisms are involved in the vascular response to ischemia/hypoxia, including the activation of inflammatory signaling pathways, oxidative stress and the production of nitric oxide [
19]. Genetic factors might also affect the risk for ROP, even though no one has been identified thus far. The disease progresses more often in white than black infants and in boys than girls [
20,
21].
The role of the β-adrenergic system
Propranolol is a non-selective β-adrenoreceptor (β-AR) antagonist. For many years, it has been largely used in the pediatric population affected by cardiovascular diseases (i.e. arterial hypertension, obstructive hypertrophic cardiomyopathy, Fallot tetralogy and arrhythmia), hyperthyroidism (i.e. neonatal thyrotoxicosis), migraine and portal hypertension with gastroesophageal varices at risk of bleeding. Propranolol is also effective and sufficiently safe in treating infantile hemangioma (IH) in childhood [
22‐
24] and the European Medicines Agency (EMA) has recently authorized the use of propranolol for life-threatening IH, at risk of ulceration or permanent deformation. The working mechanisms of propranolol in reducing proliferative IH are not completely known and include vasoconstriction, induction of epithelial cells apoptosis and inhibition of angiogenesis [
25‐
27]. The growth of IH is enhanced by pro-angiogenic factors, including VEGF and basic fibroblast growth factor (bFGF) and propranolol inhibits the growth of IH by decreasing the expression of pro-angiogenic factors and Hypoxia Inducible Factor 1 (HIF-1) induced by adrenergic receptors [
26‐
32]. Some pathogenic aspects of ROP are probably common to IH, as suggested by the evidence that ROP and IH often coexist in infants weighting <1250 g [
33] and that both diseases share the same histological feature. For instance, endothelia of IH and of retinal neovasculature in ROP express GLUT1 [
34,
35], a factor significantly up-regulated in hypoxic tissues and stimulated by HIF-1 [
36].
Additionally, as for IH, the vascular proliferative phase induced by hypoxia, which is the “second phase” in ROP pathogenesis, is promoted by VEGF. Considering that both β1 and β2-ARs are expressed in the retina [
37‐
40], that hypoxia increases VEGF levels presumably through overactivation of the β-adrenergic system as suggested by norepinephrine accumulation in response to hypoxia [
41,
42], that β-AR blockade is effective in mouse models of retinal neovascular diseases, our assumption was that the use of β-AR blockers, such as propranolol, could be useful for the treatment of ROP in infants. Indeed, several studies using a mouse model of oxygen-induced retinopathy (OIR) [
43,
44] have analyzed the role of the adrenergic system in the ROP pathogenesis and the effects of β-AR antagonists and agonists on ROP development [
45‐
47]. These studies confirmed that retinal exposure to hypoxia leads to an increase in catecholamine release, which promotes the up-regulation of pro-angiogenic factors and retinal angiogenesis by over-activating β-ARs [
46]. The β-AR blockade by systemic propranolol administration decreases VEGF and IGF-1 levels, retinal hemorrhage, retinal tufts and blood-retinal barrier breakdown, improving the retinopathy score [
45]. Similar findings were observed using selective β2-AR blockade [
47] and after β2-AR desensitization following agonist administration [
46], confirming that β2-ARs play a central role in the pathogenesis of ROP.
However, these findings obtained in C57BL/6 mice seem to conflict with results reported in 129S6 mice, a strain predisposed to develop a more aggressive neovascularization [
48] and characterized by an impressive up-regulation of β3-ARs [
49]. In this strain propranolol does not seem to affect the retinal response to hypoxia [
49], but our hypothesis was that probably the different genetic background of the mouse strain might contribute to the different retinal responses to hypoxia [
50]. The hypothesis that the insensitivity to propranolol of 129S6 mice was due to the preponderance in this strain of β3-ARs, that are minimally blocked by propranolol [
51], was confirmed by the discovery that this receptor is involved in VEGF production in hypoxic retinas, through the nitric oxide pathway [
52]. The discovery of a proangiogenic action of β3-ARs suggested to investigate a possible role for this receptor also in cancer growth [
53‐
55], a new frontier of research currently for neonatologists.
Efficacy and safety of oral propranolol
The studies in the OIR model provided a considerable amount of results which strongly indicate that β2-AR blockade may play a significant action against hypoxia-induced retinal neovascularization. This evidence prompted an interest in exploring the possibility that the administration of propranolol may not only be used to treat IH but also be of help in the treatment of ROP. A randomized controlled trial [
56] was performed to verify the efficacy and safety of oral propranolol in preterm newborns (GA < 32 weeks) with ROP stage 2 without plus in zone II [
57]. Oral propranolol significantly decreased ROP progression to both stage 3 and stage 3 with plus, and none of treated newborns progressed to stage 4. The number of newborns who underwent laser photocoagulation or bevacizumab administration was significantly lower in the treated group [
56]. These data are consistent with those reported by other authors [
58‐
60]. Despite propranolol being generally safe and well tolerated in infancy, serious adverse events have been reported in unstable preterm newborns, mainly in conjunction with other conditions, such as sepsis, anesthesia or tracheal stimulation [
56]. In these patients receiving the lower dose of 1 mg/kg/day, mean propranolol plasma concentration was around 20 ng/mL. Considering that pharmacological effects of β-blockers are usually related to the plasma concentrations, it appears prudent to avoid in future clinical trials propranolol concentrations higher than 20 ng/mL, that was considered a sort of safe cut-off value [
56]. Although propranolol is effective in counteracting ROP progression [
56,
58‐
60], the incidence of adverse events indicates that systemic administration is not sufficiently safe in preterm newborns [
56]. Recently, also prophylactic propranolol administered on seventh day of life showed a decreasing trend in the incidence of ROP, need for laser therapy, and treatment with anti-VEGF [
61].
Efficacy of propranolol eye drops in animal models
Since the oral administration of propranolol did not guarantee adequate safety, further experiments investigated the efficacy and safety of topical propranolol, in the form of eye drops, in animal models. In 2013, Dal Monte and co-workers demonstrated that 2% topical propranolol administration provides the retina with a drug concentration that are adequate to decrease pro-angiogenic factors (VEGF and IGF-1), retinal angiogenesis and blood-retinal barrier breakdown in OIR mice [
62]. The efficacy and safety of topical propranolol were also evaluated in a rabbit model [
63]. Male New Zealand white rabbits were treated with propranolol-based ocular drops at 0.1% of concentration, applied every 6 h to both eyes for 5 days. Retinal and plasma concentrations of propranolol were measured and compared with those registered after oral treatment. Despite retinal drug concentrations being similar to those reported after oral treatment, plasma propranolol levels were significantly lower after topical administration. Additionally, Draize test (a classical acute toxicity test) and cornea’s histological analysis showed no significant differences between control and treated eyes, confirming that local tolerability of ocular propranolol drops was optimal. All these findings suggested that topical propranolol formulation might be equally effective as systemic administration but have a better safety profile.
Safety and efficacy of propranolol 0.1% eye micro-drops in newborns
Recently an open-label, trial was performed to evaluate the safety and efficacy of propranolol 0.1% eye micro-drops in preterm newborns with stage 2 ROP without plus [
64]. The study was planned according to the Simon optimal two-stage design for phase II clinical trials and it was discontinued before starting the second stage since the number of failures was above the set threshold. Even though the objective to move to the second stage was not reached, the percentage of ROP progression (around 26%) was substantially similar to that obtained after oral propranolol administration. Nevertheless, no adverse effects were observed and propranolol plasma levels were significantly lower than those measured after oral administration (consistently below the cut-off value of 20 ng/mL). Therefore, treatment with propranolol 0.1% eye micro-drops seems to be safe and well tolerated in preterm newborns, but not sufficiently effective in reducing ROP progression.
Further research is then required to identify the optimal dose and schedule of topical propranolol therapy for ROP.