Article Text

Download PDFPDF

Predicting complications with pretreatment testing in infantile haemangioma treated with oral propranolol
  1. Lilian YH Tang,
  2. Jeremy WG Hing,
  3. Jasmine YM Tang,
  4. Hiroshi Nishikawa,
  5. Hossain Shahidullah,
  6. Fiona Browne,
  7. Ashish Chikermane,
  8. Manoj Parulekar
  1. Birmingham Children's Hospital, Birmingham, UK
  1. Correspondence to Dr Manoj Parulekar, Birmingham Children's Hospital, Steelhouse Lane, Birmingham B4 6NH, UK; manoj.parulekar{at}


Background Since 2008, orally administered propranolol has rapidly gained acceptance as the preferred therapy for haemangiomas, and is usually initiated by ophthalmologists, dermatologists or plastic surgeons who do not routinely use propranolol for any other indication. During the initial years when experience was limited, most healthcare professionals justifiably adopted a cautious approach when initiating and monitoring treatment. A consensus recommendation from the American Society of Dermatologists suggests routine observation, monitoring and cardiology assessments prior to propranolol initiation.

Aim This study aims to analyse treatment initiation in a large tertiary children's hospital and investigate the value of pretreatment testing in predicting commonly seen adverse reactions of propranolol.

Method 104 eligible patients treated between January 2009 and July 2012 were included. All patients underwent pretesting either with protocol A (administration of test dose with routine observations) or protocol B (cardiology clinic assessment, including two-dimensional echocardiography without test dose).

Results 38.5% (40/104) of patients developed adverse reactions during treatment; however, there were no severe or life-threatening reactions. Protocol A has a sensitivity of 0 (95% CI 0 to 0.17) and specificity of 0.95 (95% CI 0.83 to 0.99). Protocol B has a sensitivity of 0.07 (95% CI 0 to 0.34) and specificity of 0.86 (95% CI 0.63 to 0.96).

Conclusions The predictive values of both protocols for the commonly observed adverse reactions are low. In this series, there is no evidence to suggest that routine pretreatment testing before propranolol initiation is of any value in otherwise healthy children.

  • Child health (paediatrics)
  • Drugs
  • Pharmacology
  • Treatment Medical
View Full Text

Statistics from


Propranolol is a synthetic non-selective β-blocker, which has been in use for five decades, and its adverse reaction and safety profile are well documented in adults. The efficacy of propranolol in infantile haemangioma (IH) was first reported in 2008 by Léauté-Labrèze et al,1 where a child being treated with propranolol for cardiomyopathy, had coincidental regression of their nasal capillary haemangioma. Since then, it is generally accepted that propranolol has revolutionised the treatment of complicated IH.2–4 However, to date, there is only one active phase II/III Investigational New Drugs application ( NCT1056341) for the use of propranolol in treating IH.

In a recent review, the most frequently reported serious complications during propranolol therapy in IH were hypotension, pulmonary symptoms related to direct blockade of adrenergic bronchodilatation, hypoglycaemia or hypoglycaemic seizures, bradycardia and hyperkalaemia.5 Other adverse reactions include lethargy, somnolence, sleep disturbance and diarrhoea.5,6 A systematic review of 41 studies by Marqueling et al7 showed an adverse reaction rate of 31.2% (371/1189) during treatment.

Despite the increase in the use of propranolol for IH worldwide, there are no established guidelines regarding safety monitoring and dose escalation. In the UK, most major centres (as is the case with our centre) tended to follow variants of two main test protocols, as illustrated in our methodology. The evidence behind some of the practices advocated are unclear, specifically with regard to practice of test dosing or cardiac investigations, as defined in the respective protocols. To date, we are not aware of any evidence to support either test dosing or screening cardiac investigations in all children with IH treated with propranolol. The aim of this study was to assess the value of test dosing and screening cardiac investigations in predicting the likelihood of complications during treatment with propranolol in IH.


Study design

This was a retrospective cohort study carried out between July 2012 and December 2012 at a large tertiary children's hospital. The study analysed all patients who were diagnosed and treated for IH between January 2009 and July 2012.

Subjects and setting

The study was carried out at Birmingham Children's Hospital, Birmingham, UK after approval by the institutional audit committee. Inclusion and exclusion criteria are shown in table 1. Multiple databases within the centre were cross-checked to ensure maximum case ascertainment. This included the cardiology, dermatology, ear-nose-throat (ENT), ophthalmology, plastic surgery databases, medical day-care ward admission logs and pharmacy prescriptions issued for propranolol.

Table 1

Patients characteristics

Data capture

Medical records of all cases with a potential diagnosis of IH were requested and analysed by two authors (LYHT and JWGH). After excluding ineligible cases, the data collected included basic demographic statistics, date of start of treatment, location of IH, test protocol used, adverse reactions during pretreatment testing and during treatment, treatment dose strength and duration of treatment.


There were two pretreatment test protocols in use in the centre during the study period:

  • Protocol A: the patient was scheduled for a day-care ward admission for baseline observations and a single dose of propranolol 1 mg/kg body weight administered, with observations for the subsequent 2 h (blood pressure (BP) and heart rate every hour, and blood glucose at 2 h). The patient was commenced on the full dose (2 mg/kg in two divided doses), if there were no adverse reactions noted, specifically hypotension (<70 mm Hg systolic BP), bradycardia (<70 bpm) or hypoglycaemia (<3 mmol/L).

  • Protocol B: the patient was scheduled for a cardiology outpatient review and assessment, which includes baseline observations, ECG and a two-dimensional echo. If these are unremarkable, the patient was commenced on half-dose treatment (1 mg/kg in two divided doses) for 1 week, and if tolerated, doubling to the full dose thereafter.

The primary outcome of the study is to analyse the predictive values of both test protocols for adverse reactions of treatment. The secondary outcome is to document the adverse reaction profile in our series. We considered adverse reactions to be severe if they were life-threatening or if they required urgent medical attention, such as symptomatic bradycardia or hypotension, acute asthma or hypoglycaemic seizures. Due to the retrospective nature of the audit, allocations to both protocols were not planned. Instead, the use of protocol B only emerged later as clinicians became more familiar with the treatment, eliminating the need for inpatient admission.

Statistical analysis

Patients undergoing the two protocols were analysed separately. Statistical analysis was performed using VassarStats online statistical package (Richard Lowry 1998–2014). Basic descriptive statistics were performed, and predictive values along with 95% CIs were calculated. Due to small numbers in this series, two-tailed Fisher's exact test is used to calculate significance, with a p value of <0.05 considered significant.


Inclusion and exclusion

After the initial database search, there were a total of 202 patients with a potential diagnosis of IH, of which, 177 patients had a confirmed diagnosis of IH, with 22 cases from dermatology, 9 from ENT, 1 from general paediatrics, 122 from ophthalmology and 23 from plastic surgery.

After excluding patients managed without propranolol or those with incomplete follow-up or incomplete records, a total of 104 patients (68 patients in protocol A and 36 patients in protocol B) were included for analysis.

Patient characteristics

The patient characteristics are shown in table 2. The vast majority of IH occurred on the face (83/104 (79.8%)), of which, 53 (51%) patients had periocular haemangiomas. Treatment was commenced by the age of 6 months in the majority of patients (52/91 (57.1%)) and by 1 year in most patients (77/91 (84.6%)).

Table 2

Protocol A

Mean duration of follow-up was 7–9 months with no statistically significant difference between the 2 groups (protocol A: median=7.9 months (IQR, 4–24 months), protocol B: median=7.9 months (IQR, 4–12 months)).

Protocol A (test dosing)

Of the 68 patients in this group, 2 patients (2.9%) suffered adverse reactions during pretreatment testing. One child was recorded with a systolic BP of 34 mm Hg, which did not correlate with the rest of the clinic signs, and was confirmed to be a spurious reading on repeat testing. One child developed hypoglycaemia (blood glucose: 3.2 mmol/L). Both patients were commenced on full dose treatment, with specific advice about the importance of regular feeding while on propranolol therapy, and to omit a dose if the child is unwell and not feeding satisfactorily. During treatment, 25 patients (36.8%) went on to develop adverse reactions secondary to propranolol. The two patients who had adverse reactions on testing did not experience any adverse reactions on treatment.

Predictive value of protocol A

Table 2 is the contingency table comparing adverse reactions during test dosing versus treatment. In our series, test dosing does not appear to accurately predict adverse reactions during propranolol therapy (p=0.53). The sensitivity of protocol A is 0 (95% CI 0 to 0.17) and specificity is 0.95 (95% CI 0.83 to 0.99). The positive predictive value of protocol A is 0 (95% CI 0 to 0.80) and negative predictive value is 0.62 (95% CI 0.49 to 0.73).

Protocol B (cardiology assessment)

There were 36 patients included for analysis. Of the 36 patients, 4 patients (11.1%) had abnormal cardiac reviews. One patient had a non-pathological cardiac murmur, two patients had a patent foramen ovale, and one had a moderate ostium secundum ASD with left-to-right shunting. However, all these patients were cleared for propranolol therapy. Another patient had atrial flutter and impaired ventricular function as a neonate, but had a normal cardiology assessment at commencement of treatment. None of these patients developed any adverse cardiac reactions during treatment. Fifteen patients (41.7%) in this group developed adverse reactions secondary to propranolol.

Predictive value of protocol B

In our series, cardiology assessment does not appear to accurately predict adverse reactions during propranolol therapy (p=0.62) (table 3). For any adverse reaction, the sensitivity of protocol B is 0.07 (95% CI 0 to 0.34) and specificity is 0.86 (95% CI 0.63 to 0.96). The positive predictive value of protocol B is 0.25 (95% CI 0.01 to 0.78) and negative predictive value is 0.56 (95% CI 0.38 to 0.73).

Table 3

Protocol B

For adverse cardiac reactions alone, defined as bradycardia and/or hypotension (table 4), the sensitivity of protocol B is 0 (95% CI 0 to 1) and specificity is 0.89 (95% CI 0.73 to 0.96). The positive predictive value of protocol B is 0 (95% CI 0 to 0.60), and negative predictive value is 1 (95% CI 0.87 to 1).

Table 4

Protocol B

Adverse reaction profile of propranolol

A significant proportion of patients (40/104 (38.5%)) suffered from some form of adverse reactions during propranolol therapy, with 8 (7.7%) of these patients having multiple adverse reactions. The majority of adverse reactions were breathing-related problems (12/104 (11.5%)), cold extremities (10/104 (9.6%)), gastrointestinal (watery diarrhoea) (10/104 (9.6%)) and sleep-related changes (night terrors) (9/104 (8.7%)) (figure 1).

Figure 1

Adverse effects of propranolol therapy.

In our series, there were no patients who suffered any severe or life-threatening adverse reactions to propranolol therapy. The adverse reaction profile in our series appear to be in keeping with the adverse reaction profile as reported by Marqueling et al.7

Patients who experienced breathing difficulties during propranolol therapy had a mean age of 6.2 months at commencement of treatment (range, 3.1–9.4 months). These were reported by the patient's parents as wheezing and mild nocturnal distress. None of these were respiratory emergencies requiring urgent admission or intervention. Of these 12 patients, two, who were found evidently wheezing and distressed, were maintained on half-dose propranolol therapy rather than escalation to the full dose. The remaining patients were well at the time of examination.


The study cohort was typical for IH, with the majority of patients under 1 year age at commencement of treatment.

Over one-third of patients suffered some adverse reaction during propranolol therapy, and the incidence of adverse reactions was as expected in this study cohort, consistent with other reported series. There were no serious adverse reactions, and alteration of dose was successful in resolving the adverse reactions in the majority of cases. The most commonly encountered adverse reactions were breathing-related problems, cold extremities, gastrointestinal upset and sleep disturbance, in approximately 10% of patients each.

Where patients were admitted to a day-care ward to receive a test dose of propranolol, we observed no significant association of adverse reactions between test dosing and treatment. For those patients who received cardiology assessment, again, there was no significant association between positive investigations and the likelihood of developing adverse reactions during treatment.

There are several recent reports regarding cardiac pretesting prior to propranolol treatment for IH. Drolet et al5 reported that while routine ECG screening is advocated, its utility in all children with IH before treatment with propranolol is unclear. Raphael et al8 concluded that pretreatment ECG is of limited value for patients with an unremarkable cardiovascular history and a normal heart rate and BP, and that hypotension may occur during treatment. Blei et al9 found a high incidence (21%) of coexistent echocardiographic cardiac anomalies in their cohort, none of which precluded the use of propranolol.

Therefore, it is likely that a more indication-driven ECG strategy will be developed, as the incidence of ECG abnormalities that would limit the use of propranolol in IH is low.5 In this same consensus report, the authors have not advocated the use of echocardiography as a routine screening tool in the absence of abnormal clinical findings. Several papers have also reported adopting a more indication-driven approach in using ECG10 and echocardiography3 ,10 for screening in children with signs of a syndromal IH associated with cardiac pathology, such as PHACES (posterior fossa malformations–haemangiomas–arterial anomalies–cardiac defects–eye abnormalities–sternal cleft and supraumbilical raphe syndrome), cardiac murmur, ECG abnormalities, or signs of cardiac failure and in children with a large bulky IH.3

This study does not demonstrate any evidence to suggest that pretreatment testing, with propranolol test dose or cardiac investigations, is effective in predicting adverse reactions of oral propranolol treatment for IHs. It is important to educate parents about the potential adverse reactions of oral propranolol treatment and provide easy access to care when indicated. Parents should also be advised to withhold one or more doses of propranolol if the child is not feeding to avoid hypoglycaemia or in the event of breathing difficulties and seek urgent physician review.


There is no evidence currently to suggest that routine pretreatment testing on otherwise healthy children with IH treated with propranolol is of any value. However, we would recommend further studies and investigations to develop a more indication-driven strategy for pretreatment cardiac investigation in suitable cases.


View Abstract


  • Contributors LYHT, JWGH and MP designed the study. LYHT, JWGH, JYMT and MP performed the literature review. LYHT and JWGH collected the data. JWGH performed statistical analyses. MP supervised the study. All authors contributed to the manuscript.

  • Competing interests None declared.

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

  • Data sharing statement All authors have full access to all data, including statistical reports and tables in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis.

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.

Linked Articles

  • At a glance
    Keith Barton Jost Jonas James Chodosh