Risk of neurodevelopmental abnormalities in children treated with propranolol

Dear Editor, Since the discovery of the clinical benefit of propranolol in infantile haemangioma,1 a substantial amount of data has been accumulated leading to its recent proof of efficacy2 and approval by health authorities for this indication. Especially in young children the possibility of long‐term behavioural consequences of propranolol use has been a constant concern, but there have been no data other than purely theoretical considerations to support this.3 Even if reporting of memory disorders or neurodevelopmental defects by cardiologists has been underestimated, these side‐effects have never been reported in over 40 years of use. It cannot be denied that propranolol is a lipophilic molecule that crosses the blood–brain barrier and has some central nervous system effects. Most are unwanted, such as the sleep disorders observed in 22% of patients treated with propranolol (3 mg kg−1 per day for 6 months) compared with 13% with placebo;2 others are expected and utilized to treat certain conditions such as post‐traumatic stress disorder (PTSD). In studies enrolling small samples (mostly < 20) of healthy adult volunteers, propranolol has been found to impair short‐ and long‐term memory, psychomotor function, sleep quality and mood. However, as alluded to by the authors themselves and as advised by specialists in psychology,4 the statistical power of the cited neurological studies should be weighed up before drawing definitive conclusions. In addition, application of adult observations to children is at best uncertain, and it is questionable whether parallels can be drawn between decreasing the emotional intensity of the traumatic memory in PTSD in adults and normal development in infants. As lead investigator for the clinical trials conducted leading to the approval of propranolol solution in infantile haemangioma, I can bring objective data to this discussion. A study has been performed using a well‐established juvenile toxicity model to test the effects of propranolol on development in juvenile rats, covering the developmental periods corresponding to infancy (treatment period), childhood and adolescence. Three groups of 64 juvenile Sprague–Dawley rats (50 : 50 male : female) received daily oral propranolol (by gavage) from postnatal day 4 (PND 4) to PND 21 at 10, 20 or 40 mg kg−1 per day; a control group received vehicle alone under identical conditions. During the time corresponding to the peripubertal period there were no product‐related findings on functional observation battery reactivity; motor activity measurement showed no relevant differences from controls; and there were no indications of product‐related impairment of learning or memory assessed by the water multiple T‐maze test (personal communication, as included in the nondisclosure agreement). The recently published study2 is the only large‐scale randomized controlled trial with a significant (2‐year) follow‐up. Among 322 patients treated with propranolol for 24 weeks, seven (2%) had an abnormal neurodevelopmental assessment at week 96, comprising abnormalities present at inclusion (two), isolated strabismus (two), delayed neurodevelopment not confirmed by a developmental paediatrician (two) and mild delays (two). Comparison with the placebo group was not possible. A noncontrolled prospective study in infantile haemangioma5 reported 13 of 200 propranolol‐treated patients (6·5%) with gross motor abnormalities including delayed walking. A retrospective survey of patients prescribed propranolol for infantile haemangioma reported four of 84 patients (5%) with delayed walking unassisted.5 In both studies all infants were walking by 20 months, and in one a statistically significant influence was observed for patients taking other medications.6 As infants with severe haemangiomas, and particularly those with PHACE syndrome, often present gross motor and speech delay,7 the neuropsychological impact of treatments vs. disease is difficult to establish. The neurodevelopmental defect/delay rates observed with propranolol seem similar to those observed in the general population. The Centers for Disease Control and Prevention report as many as one in six children having a developmental disorder, including 3·7% with developmental delays.8 A prospective study of 348 newborns9 observed abnormal development at 24 months in 29 (8·5%), including global developmental delay in five (1·5%) and behavioural/sleep disorders in 12 (3·5%). The prevalence of developmental delay has been estimated at 3·3% in children aged < 5 years, including deficits in gross motor function (0·9%) and overall delay (1·1%).10 In conclusion, before initiating systemic treatment with any beta‐blocker in severe infantile haemangioma, it should be reminded that, with thousands of published cases, > 1600 patients well documented in a compassionate‐use programme and 435 patients thoroughly monitored in clinical trials, propranolol is the beta‐blocker with by far the best‐documented safety profile, even if possible long‐term effects remain to be evaluated. To avoid bias when assessing the effects of any beta‐blocker on the memory process, future studies should be well powered, take into account childhood amnesia [it has been demonstrated that adults retain few memories of early childhood (average earliest memory is from 3–4 years old); this important and positive process, known as ‘childhood amnesia’, permits ‘forgetting’ of early childhood memories to allow the development of a concrete, complex, autobiographical memory]11 and carefully control for age, treatment and infantile haemangioma condition – essential study design aspects required to answer this question adequately. Funding sources: none. Conflicts of interest: C.L. is a consultant for Pierre Fabre Dermatologie, and J.J.V. and A.D. are employed by Pierre Fabre Dermatologie.