Article Text

Download PDFPDF

The use of propranolol in the treatment of periocular infantile haemangiomas: a review
  1. I Claerhout1,
  2. M Buijsrogge1,
  3. P Delbeke1,
  4. S Walraedt1,
  5. S De Schepper2,
  6. B De Moerloose3,
  7. K De Groote3,
  8. C Decock1
  1. 1Department of Ophthalmology, Ghent University Hospital, Belgium
  2. 2Department of Dermatology, Ghent University Hospital, Belgium
  3. 3Department of Pediatrics, Ghent University Hospital, Belgium
  1. Correspondence to Dr Ilse Claerhout, Department of Ophthalmology, Ghent University Hospital, De Pintelaan 185, Gent 9000, Belgium; ilse.claerhout{at}


Infantile haemangiomas (IH) are benign vascular tumours characterised by their very rapid growth. Although usually innocuous, periocular IH can cause serious visual loss through induction of strabismic, deprivational or anisometropic astigmatism. Common treatment modalities for these IH include intralesional and systemic oral steroids; however, both treatments are associated with potentially severe side effects. A report was published recently demonstrating the impressive effect of propranolol in the treatment of IH. This exciting finding has provoked a paradigm shift in the management of this condition. So far little has been reported in the specific ophthalmologic literature, although case reports are emerging. This review gives an overview of the recent findings and includes the authors' experience with 10 patients treated with propranol.

  • Propranolol
  • infantile haemangioma
  • safety protocol
  • eye lids
  • neoplasia
  • drugs
  • treatment medical
View Full Text

Statistics from


Infantile capillary haemangiomas (IH) are benign vascular tumours that have a typical clinical course, characterised by a rapid proliferative phase in early infancy followed by an involutional phase. IH affects one in every 10 to 20 Caucasian infants. It is more common in females with a 3:1 to 5:1 proportion, and has a higher incidence in premature infants, in newborns with low birth weight, in those infants whose mother's pregnancy presented placentary anomalies, and in neonates from mothers with multiple pregnancies. Fifty to sixty per cent of IH affect the head and neck area.1

IH are the most common tumours of the eyelid and orbit in infants. They typically present as small tumours at birth or during the first 2–3 months of life.2 Their rapid growth is the historical hallmark of IH. The proliferation phase occurs during the first year, with the most prominent growth occurring during the first 4–6 months of life. Proliferation slows considerably between the middle and end of the first year of life. During this time, the haemangioma may remain quiescent or may begin to involute.3 The involutional phase occurs over a period of 5 to 7 years. As a rule of thumb, 50% of IH complete involution by age 5 years and 70% by age 7 years. The remainder may take an additional 3–5 years to complete the process. Even after involution, permanent cutaneous residua may be present under the form of scar formation, telangiectasia or redundant skin.4

Despite their self-limited course, IH can impair visual function and for that reason are a frequent cause for referral to the ophthalmologist. The incidence of ophthalmic complications has been reported to be as high as 46% to 80%.5–7 Tumour growth within the eyelid and orbit can lead to ptosis with obstruction of the visual axis and deprivational amblyopia, and to translational displacement of the globe resulting in strabismic amblyopia. Tumour growth can also induce mechanical deformation of the structurally immature cornea causing anisometropic astigmatism and refractive amblyopia.8 A retrospective review9 of 129 patients with periocular IH suggested that size greater than 1 cm in largest diameter is an important predictor of amblyogenic factors and approximately half of these patients will require treatment.

To prevent visual loss from either occlusion or anisometropia, several treatment modalities have been shown to be successful in shrinking the size of the IH. A recent report10 reviewed the various pharmacological agents in the treatment of periocular IH, including steroids (systemic or intralesional), interferon, vincristine, cyclophosphamide or imiquimod. They concluded that there is no single uniformly safe and effective treatment able to sufficiently control development and growth. Intralesional steroid injections have long been the treatment of choice for periocular IH.11 Nonetheless a growing list of local complications and recognition of systemic complications have raised concerns and led to a shift to treatment with systemic corticosteroids.12 13 Given the side effects of systemic corticosteroids, there is an unsatisfied medical need for a first-line treatment with a good benefit–risk profile. Recently, the use of propranolol (a beta blocker) was serendipitously found to reduce the size of IH.14 This review will focus on the published current evidence of its efficacy and safety and includes the authors' experience with 10 patients treated with propranol.

First evidence on the efficacy of propranolol

In 2008 Léauté-Labrèze et al14 reported on their observation that propranolol can inhibit the growth of IH. The observed effect of propranolol was serendipitously discovered with an attempt to treat the adverse cardiac effects that developed as a result of high-dose systemic corticosteroids in two children with severe haemangiomas (one with a nasal IH and the other with an IH involving the entire upper limb, part of the face and extending intracervically). In both children the positive effect of propranolol at a dose of 2 mg/kg per day on involution of the IH was quite dramatic and very rapid (1 and 7 days, respectively) with stable results after the end of treatment, which suggested evidence of the benefits of this drug. Both children also concomitantly received systemic steroids.

After observing the success in these two cases, propranolol was given to nine additional children who had severe or disfiguring haemangiomas. In all patients, 24 h after the initiation of treatment, there was a change in the haemangioma colour from intense red to purple, and this was associated with a palpable softening of the lesion. This very favourable outcome prompted many physicians managing this disease to use propranolol as a first-line treatment.

Mechanism of action

Little is known on the exact working mechanism of this non-selective beta blocker in the treatment of IH. One possible explanation for the effect of propranolol is induction of vasoconstriction, which is immediately visible as a change in colour, associated with a palpable softening of the haemangioma.14

IH consist of a complex mixture of cell types. Immature endothelial cells coexist with immature pericytes and different angiogenic peptides, such as basic fibroblast growth factor (FGF) and vascular endothelial growth factor (VEGF), induce proliferation of these immature cells, resulting in the development of the haemangioma.15 Beta blockers could potentially influence the signal transduction pathway of these angiogenic factors through modulation of the beta-adrenergic receptor system16, thus explaining their effect in the proliferative phase. Other studies demonstrate that non-specific beta blockers such as propranolol are able to trigger apoptosis in capillary endothelial cells in adult rat lung tissue.17 A similar mechanism might be applicable to haemangioma endothelial cells, resulting in an induction of involution.

Protocol for initiating propranolol therapy

The seminal paper14 on the use of propranolol for IH did not elaborate on details for initiating therapy, monitoring and potential risks. Soon after, two publications18 19 described the potential risks of this drug and made recommendations for a treatment protocol to minimise potential adverse events.

Propranolol is a non-selective beta blocker used mainly in the treatment of hypertension. It has also been used in paediatric cardiology to treat a number of congenital cardiac anomalies and arrhythmias.10 The most common side effects of beta blockers are well known: these include transient bradycardia and hypotension that warrant close monitoring at the onset of treatment. Bronchospasm is usually only seen in patients with an underlying sensitivity. Therefore a family history regarding atopy and asthma and a specific history taking for repeated wheezing in the child should be performed. Propranolol may also blunt the clinical features of hypoglycaemia; however, in children past the neonatal period of the first week of life, when spontaneous hypoglycaemia is less likely to develop, this does not seem to be much of a concern. Its use should be avoided during the first week of life, but overall, however, oral propranolol is considered a safe and well-tolerated medication even in children.20

There is no universally accepted protocol for initiating propranolol therapy, but it seems wise to follow some guidelines and to perform a baseline electrocardiogram, with frequent monitoring of vital signs (blood pressure and pulse) when first starting the medication. Some authors advocate admitting the children to hospital;19 however, a recent report by Haider et al21 demonstrated a good safety profile when starting propranolol as an outpatient treatment. This does require careful observation by care givers and a close interaction with the primary care physician.

Some authors advocate a gradual increase in loading dose,19 21 22 whereas others14 23 start at a dose of 2 mg/kg per day straightaway.

Special care should be taken in children with very large segmental haemangiomas that cover a particular section or area of skin, as these may be markers for underlying malformations or developmental anomalies of the heart, blood vessels or nervous system (the so-called PHACES24 and PELVIS25 syndromes). These children are at higher risk for cerebral vascular accidents secondary to cerebral vascular anomalies and should theoretically not receive beta blockers, although Haider et al have reported good clinical results in two patients with the PHACES syndrome.21

Based on our own (unpublished) results in several patients treated with propranolol for IH, we propose a suggestion for an optimal treatment schedule. The study complied with the Declaration of Helsinki and necessary ethics committee approval was secured for the study reported. The protocol at our institution includes a thorough personal and familial history, an elaborated clinical examination by a paediatrician, an ECG, a blood pressure check and a blood glucose level test. When there are no contra-indications, propranolol is initiated at a dose of 1mg/kg per day in two divided doses. After 10–14 days, the patient has a second ECG and check of blood pressure, pulse and glucose level. If these are normal, the dose of propranolol is increased to 2 mg/kg per day in two divided doses. At home parents should observe for signs of lethargy, poor feeding and/or bronchospasm. Further follow-up consists of monthly clinical and photographic evaluations of the IH, and monitoring of treatment compliance and tolerance (pulse and blood pressure). The very first patients with IH were treated as day patients with frequent monitoring of blood pressure and pulse. However, based on the excellent safety profile obtained in these patients and on the experience in paediatric cardiologic patients who receive much higher doses of propanolol, patients are currently started on therapy on an outpatient basis. None of our patients with periocular IH had a contra-indication to starting treatment with propranolol. In our much larger cohort of non-periocular IH we had one child who could not start treatment because of wheezing.

Clinical results in periocular IH

Because of the recent introduction of propranolol in the treatment of IH, initial case reports often report on its success in conjunction with or following unsuccessful steroid treatment. This was also the case in the two index cases.14 Currently, however, more case series are being published on the experience of using propranolol as the sole treatment for IH in different subspecialties.22 26 27

Despite the relatively high output of papers in the last 2 years on propranolol and haemangioma (50 hits on Pubmed in June 2010), the published reports on specific periocular cases in the ophthalmologic literature and journals remain scarce.

Taban and Goldberg28 published the case of a 6-week-old girl with a large refractory left orbital and periorbital haemangioma. Despite intralesional steroid injection and systemic oral steroids, the tumour had not responded and had slightly progressed. When propranolol treatment at a dose of 2 mg/kg per day was started (in conjunction with steroids), she showed dramatic improvement, starting days after the initiation of the treatment. Simultaneously, steroids were tapered over 2 weeks. Propranolol was continued for 2 months without regrowth of the lesion after stopping therapy.

Fay et al29 reported on a case in which propranolol was used as the single initial treatment modality in a case of a 4-month-old girl with an intraconal IH inducing protrusion of the right eye. Propranolol was started at a dose of 2 mg/kg per day and at the 6-week follow-up visit there was no obvious proptosis. Repeated MRI at the 3-month visit revealed complete resolution of the tumour. Treatment with propranolol was continued for 8 months and then tapered over a 1-month period.

Apart from these two case reports, only two cases series have been published at the time of writing this review. The largest comprises 17 patients, published by Haider et al21 After the initial case report by Léauté-Labrèze et al, Sans et al23 reported on the follow-up of 32 children including 13 patients with a periocular location of their IH. We have added the unpublished results on our own case series of 10 patients for comparison. Details of the different case series are summarised in table 1.

Table 1

Details on the case series reported in the literature, including the unpublished findings of our group, on the treatment of infantile haemangiomas with oral propranolol

If one defines success as stopping growth or reducing size, all series report a 100% success rate. These data strongly support the current perception that propranolol is a highly successful first line treatment modality with very limited and mild side effects. Caution is warranted as ‘stopping growth’ is a rather difficult end point in a disease with such a variable clinical course as IH. The same applies to ‘reduction in size’ given the fact that most IH will involute spontaneously. Sans et al23 report an efficacy of 100% with growth stabilisation and improvement continuing until complete involution in all their cases. Haider et al21 described the response to therapy as being excellent (>50% reduction in size), good (decreased size but <50%), fair (no further growth) or poor (continued growth or intolerable adverse effects). Based on this classification, 10 of their patients (59%) had excellent results, six (35%) had a good response, one (6%) a fair response and none a poor response. They reported only one patient with complete resolution.

Our unpublished results are described using the same classification. Figure 1 shows two children before and after treatment with propranolol. For evaluating the response to therapy we averaged the score of three blinded independent observers. We had five patients (50%) with excellent results, three (30%) with a good response and two (20%) with a fair response. In the two children with a fair response, the IH did not show any reduction in size, but did stop growing. We also had one complete resolution. One of the patients with an excellent result and nearly complete resolution had to stop treatment because of bronchoconstriction in conjunction with a lower respiratory tract infection. Our results are similar to those of Haider et al21 but in no way equal to the 100% efficacy reported by Sans et al.23 An important anecdotic finding in our series is that of an increase in size after abruptly and unplanned stopping of the propranolol. The same was published in the report of Haider et al. This supports the idea that propranolol is indeed ‘active’ in the cessation of growth and it is an argument in favour of weaning off treatment.

Figure 1

Two children before ((A) and (C)) and after ((B) and (D)), respectively, treatment with propranolol for infantile haemangiomas.

In periocular IH size and growth are only one part of the equation. Objective measures of astigmatism and anisometropia are important but notoriously difficult to obtain. In a systematic review of 81 case series, Ranchod et al30 identified 25 patients from three studies for whom refractive data were reported. In the series reported by Haider et al21 seven out of eight patients with anisometropic astigmatism had resolution of the anisometropia following treatment. In our unpublished series we obtained data on astigmatism in six patients. Of these six patients five showed a reduction in astigmatism (measured before start of treatment and at end of follow-up) from 3.8 DC to 2.25 DC (p=0.018). Sans et al23 reported success in reducing astigmatism or amblyopia in all of their patients. A recent report by Weiss et al8 stressed astigmatism and not anisometropia or amblyopia as the immediate indication for treatment of IH. Future prospective randomised studies will also have to take these endpoints into account.


Since the first report14 on the use of propranolol in the treatment of IH, a paradigm shift in the management of this disease has occurred. Several case reports and case series have been written, supporting the initial conclusions of this seminal study. However, several important questions remain unanswered, such as the ideal time to start treatment, optimal dose range, duration of treatment, resolution rate and long-term follow-up after cessation of treatment. In order to support the efficacy of treatment objective ophthalmological endpoints need to be defined. At present there are seven clinical studies registered with Clinical, six of which are already recruiting patients ( Some of the trials compare propranolol with placebo, some compare propranolol with steroids, and others examine the effects of steroids used alone compared with steroids used in conjunction with propranolol. Undoubtedly these studies will be able to answer some of the remaining questions.

Meanwhile, it is important that the treatment of infantile capillary haemangiomas with propranolol gets a wider recognition in the ophthalmological community given the relatively high incidence of this disease. This is a disease that every comprehensive ophthalmologist will encounter several times in his career. Knowledge of this novel treatment modality as well as a good understanding of the need of a careful work-up of the patient before treatment is imperative.


We thank Dr WA Van den Bosch and Dr A Shah for their help in scoring the photographical evolution of the IH patients.


View Abstract


  • Funding None.

  • Competing interests None to declare.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the Ethics Committee of the Ghent University Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.