Propranolol for Infantile Haemangiomas: Certain Chances, Potential Risks

P.H. Hoeger


The British Journal of Dermatology. 2015;172(1):3-4. 

Within a few years of the serendipitous discovery of its remarkable effects on proliferating infantile haemangiomas (IH) in 2008,[1] the nonselective beta-adrenergic antagonist propranolol has become the first-line therapy for complicated haemangiomas worldwide. As evidenced in large case series and meta-analyses,[2–4] the response rate to oral propranolol (2–3 mg kg−1 per day for 6 months) is 96–98%; reported side-effects were reversible and mostly benign. Propranolol has thus virtually replaced oral corticosteroids, laser surgery or interferon and vincristine, which in former years (and for decades) were used to treat infants with obstructive, ulcerative or disfiguring IH. Unlike 'new' agents, propranolol is a generic, off-patent and readily available drug, and was used for the treatment of IH at the discretion of the prescribing physician long before randomized controlled studies[5] or official recommendations[6] were published. Its mode of action, however, is incompletely understood so far.

In this issue of BJD, two articles shed new light on the effects of propranolol. Ji et al.[7] summarize current knowledge about mechanisms potentially underlying the antiproliferative effect of propranolol in IH. Beta-adrenergic blockade targets endothelial cells, pericytes and haemangioma stem cells, initially resulting in vasoconstriction, followed by long-term tumour regression that is likely due to inhibition of both vasculogenesis and angiogenesis. This is a result of suppression of de novo growth of blood vessels from stem cells and interference with proliferation of existing vessels.

However, inhibition of angiogenesis, desirable as it is for growing vascular tumours, can be a double-edged sword. Shortly after the description of the antiangiogenic effects of interferon-alfa in young infants with life-threatening IH,[8] it turned out that up to 25% of those treated developed spastic diplegia or other motor disturbances; this effect was confined to infants and not observed in older children treated with interferon.[9]

Is there any evidence of a similar effect in infants treated with propranolol? No, and unlike interferon, it is not to be expected for several reasons. Firstly, several thousand young infants have been treated with propranolol since 2008, and any major untoward neurological effects would have been picked up even without a formal follow-up by now. Secondly, the combination of several different mechanisms of action, as outlined by Ji et al., might be more a 'physiological' mode of action – mimicking processes involved in spontaneous regression of IH – than one overriding pathway alone. In this context, it is paradoxically reassuring that the recurrence rate of IH following discontinuation of propranolol is high (15–20%),[2,4,10] implying that its antiangiogenic effects are reversible as long as spontaneous or induced regression is incomplete.

Concerns regarding potentially relevant neurodevelopmental or cognitive side-effects of propranolol, reviewed by Langley and Pope in this issue of BJD,[11] are less easily dismissible. Lipophilic beta-blockers readily cross the blood–brain barrier, and their degree of lipophilicity correlates with the incidence of central nervous system side-effects such as sleep disturbance and interference with memory.[12] Sleep disturbance, somnolence and irritability can be observed in 15–25% of all infants treated with the highly lipophilic propranolol.[2–4] In adults, propranolol has been shown to decrease certain memory functions,[13] an effect that can be exploited therapeutically in the treatment of post-traumatic stress disorders.[14] It thus appears conceivable that blockage of neural pathways critical for learning and memory could be an unrecognized (and not easily cognizable) long-term side-effect of propranolol in infants.

This deeply disturbing issue can be clarified (and, ideally, settled for good) only by meticulously designed, prospective, controlled studies addressing neurocognitive capabilities (e.g. language and memory skills) in propranolol-treated vs. untreated children. Meanwhile, preliminary studies indicate that hydrophilic beta-blockers such as nadolol[15] and atenolol[16] seem to be at least as effective against IH as propranolol. They might in the end turn out to be an even better alternative.