Oral Propranolol in Infantile Hemangiomas: Analysis of Factors that Affect the Outcome.

AIM: The primary treatment for the subset of infantile hemangiomas (IHs) which develops complication is pharmacological intervention, and propranolol has become a popular choice. Here, we evaluated the efficacy and safety of propranolol in a clinical cohort of IHs and analyzed clinical characteristics associated with a good outcome.
MATERIALS AND METHODS: We retrospectively reviewed a total of 52 IHs patients, between ages 1 and 48 months (median age: 7.5 months), who were treated with oral propranolol, with dose ranging from 2 to 3 mg/kg/day. Efficacy was evaluated using mean percentage reduction, visual analog scale (VAS), and parental satisfaction levels at week 2 and months 1, 2, 6, and 12. The adverse effects were noted and responses after 6 months were graded. Statistical analyses of the outcome were also performed for the responses with regard to age at propranolol initiation, site of lesion, and mean duration of treatment.
RESULTS: A therapeutic response with at least 50% mean percentage reduction in size was noted in 84.6% at the end of 6 months. VAS score and parental satisfaction levels correlated well with mean percentage reduction (63.7 ± 15.6) at 6 months. Patients aged <6 months and those with cephalic lesions exhibited a greater therapeutic response rate with shorter overall mean duration of the treatment.
CONCLUSIONS: Oral propranolol at 2-3 mg/kg/day dosing has shown to be effective and safe for IHs in pediatric age group. Intervention in the early proliferative phase, with especially, the cephalic lesions result in better resolution rates with shorter duration of overall treatment.

INTRODUCTION

Infantile hemangiomas (IHs) are the most common benign tumor in pediatric patients, and 60% of IHs occur on the head and neck.[1] IHs usually manifest during the 1st month of life and the proliferative phase extends for the first 2 years of age.[2] These hemangiomas may be disfiguring but are not usually life-threatening or function impairing, and most clinicians emphasize an approach of observation before any intervention. Most IHs involute spontaneously in the first decade of life after the 2nd year. However, a subset of IHs does not involute spontaneously and this may lead to serious complications and cosmetic disfigurement, and thereby functional and psychological effects on parents and the affected children.[3] Although several treatment modalities have been used in the past; propranolol, a β-blocker, has been proven to be very effective and safe for IHs therapy, and has replaced corticosteroids as the first-line treatment for IHs.[4] The mechanism of action of propranolol is not established yet but may involve microvascular vasoconstriction and modulation of angiotensin II.[5] Alterations in the cell signaling of angiogenic factors and early apoptosis of endothelial cells may also be involved.[6] In this study, our objective is to study the efficacy of propranolol in different subsets of children in our population with IHs and to analyze factors affecting the clinical outcome.

MATERIALS AND METHODS

Patients

A total of 58 consecutive patients with cutaneous IHs who were considered for propranolol therapy at our tertiary referral institution between September 2014 and August 2017 were included in this study. The diagnosis of IHs was determined by the characteristic clinical appearance and sonological findings. The indications for treatment included rapid growth, functional impairment, ulceration, infection, parental concerns, and esthetic complications. This observational prospective interpretation of data was approved by the Institutional Review Board and informed consent was obtained from parents to start propranolol therapy and to acquire serial photographs to monitor the resolution of the lesion. The study followed the tenets of the Declaration of Helsinki for research involving human subjects.

Before the initiation of propranolol treatment, a cardiac screening was done clinically by a pediatrician (with further workups such as baseline electrocardiogram and echocardiography, if suspected) to exclude any congenital cardiac disorders. The patients who were suspected to have reactive airway disease, asthma, cardiac disease, arrhythmia, Kasabach–Merritt syndrome, PHACE syndrome, diabetes, hypersensitivity, evidence of previous treatment, and hypoglycemia were excluded. A Doppler ultrasound was performed to exclude visceral hemangioma in all patients and locally for deep-seated IHs. Complete blood count, coagulation parameters, renal and liver function tests, electrolytes and blood glucose were also routinely performed.

A standardized database recorded the patient details, which included patient gender, age at treatment initiation, location, color and dimensions of the lesion, symptoms if any, antenatal history, associated anomalies/syndromes, medication and dose, clinical response (measurements at 1 month, 2 months and 6 months), and side effects/complications resulting from treatment and parental satisfaction.

Treatment regimen

Infants below 6 months were monitored as inpatients for 24 h in order to start the treatment under medical control. For the remaining patients, treatment was carried out on an outpatient basis, and the intended duration of the treatment was for at least 6 months. The parents were educated to recognize side effects and conditions that could require interruption of the treatment. Propranolol was started at a dose of 1 mg/kg/day in two divided doses for the first 24 h and then it was increased to 2 mg/kg/day if tolerated. If response was as expected and in the absence of adverse reactions, the dose was maintained for the next 6 months at least. If no response was seen after 2 months, a dose escalation to 3 mg/kg/day was attempted and continued if tolerated. In the last month of treatment, the propranolol dose was tapered by ½ for 2 weeks. If no rebound occurred, tapering continued by ½ next 2 weeks before being discontinued completely. As an exception, when the drug was started during infancy it was not ceased in the 1st year of life regardless, to prevent recurrence or rebound. Patient's body weight was checked every month and propranolol dosage was adjusted accordingly. Follow-up visits were arranged at 2 weeks, 1 month, 2 months, 6 months, and 1 year after the initiation of therapy. Withdrawal of treatment between 2 and 6 months was accepted if the lesion had completely involuted, if there was no response at all, if the patient experienced drug-related issues, or if there was a parental request to withdraw from the study. Blood pressure, heart rate, blood glucose levels, side effects, and response to treatment were documented during each clinic visit. Since the end-point of treatment was difficult to define, the following guideline was employed on individual basis; propranolol was ceased after 6 months, if the lesion had completely regressed or when the lesion remained stable for longer than 2 months. If the response was ongoing after 6 months, the treatment continued till plateau or complete resolution.

Evaluation of response to propranolol

Treatment response was assessed according to our own assessment using changes of dimensions (length and breadth), color, and contour of IH. The dimensions were measured at baseline and at each visit by one designated reviewer. The mean percentage reduction in size was determined on follow-up at 1 month, 2 months, and 6 months. Changes in size and color of the IHs were documented by clinical photographs. A visual scale to assess severity and resolution of IHs based on serial clinical images taken at baseline and on follow-up visits were also used. The investigator documented changes in color and size of the lesions on a visual analog scale (VAS) ranging from −10 to positive (+) 10 by comparing serial photographs to the baseline image. On this scale, 0 represented the baseline photograph (pretreatment), a decrease in color or size resulted in a negative (−) number, an increase in color or size in a + number. In cases of multiple lesions, each was evaluated separately, and the average was calculated. Parental satisfaction at the end of 6 months was also noted in grades (very satisfied, satisfied, and unsatisfied).

Statistical analysis

Statistical analysis was done using SPSS (IBM Corp. IBM SPSS Statistics for Windows, Version 16.0. Armonk, NY, USA). Qualitative data were expressed as frequency and percentage. Continuous data were presented as mean and standard deviation if data were normally distributed and as median and interquartile range (IQR) if data were not normally distributed. Comparison between the sites was done by Student's t-test or Chi-square test depending on the type of the data. All tests were two-tailed, and P values were used to analyze categorical data. The significance level was set at 5%.

RESULTS

Fifty-eight patients with IHs who were considered for propranolol therapy were enrolled in the study initially. Three patients were excluded as they were suspected to have contraindications for propranolol therapy (vide above) and two were excluded due to poor compliance with the medication. Yet, another patient was excluded, as the lesion was deep seated and determination of response was difficult. The remaining 52 patients were then included in the data analysis. The patients were between 1 and 48 months with a median age of 7.5 months and IQR of 7.48% (n = 25) were males. About 31 (59.6%) patients had head and neck as site of IHs and 21 (40.4%) patients had primary lesions at other sites. Five patients were born preterm, but had completed 40 weeks at the start of therapy. No patient included in the analysis received any form of treatment prior to referral, except in six cases that were treated for ulceration (with antibiotics and/or wound dressings) of IHs. About 63.5% (n = 33) was noted to have IH by 1st month of life. Five patients (9.6%) had multiple site involvement at presentation, and 8 (15.4%) had prior history of mild-to-severe infection in the hemangioma. Six patients (11.5%) had + family history and in two cases (3.8%), the lesion was detected on antenatal screening ultrasound. There were no multiple gestations in this sample. All patients were treated according to the abovementioned regimen. The mean treatment duration was 11.2 months with a range of 2–16 months. At least 6 months therapy was ensured in most patients. The treatment was withdrawn before 6 months in five patients; as the lesion completely involuted (n = 1), there was no response at all (n = 2), and there was a parental request to withdraw from the study (n = 2).

Overall, at the 1st-month visit, the percentage reduction was between 5 and 35 with the mean of 11.3 ± 6.3. By 2 months, the percentage reduction was between 10 and 60 (mean, 30.4 ± 10.3) and after 6 months, it was between 25 and 95 (mean, 63.7 ± 15.6). These figures matched well with the VAS improvement at 6 months (37.2% with VAS scores −9 and −10, 49.1% with VAS scores between −5 and −8, 10.7% between −2 and −4, and 3% with scores <−2). Parental satisfaction was documented as very satisfied in 36 (69.2%) patients, satisfied in 13 (25%), and unsatisfied in 3 (5.7%).

Achauer et al.[7] graded the hemangiomas as: Grade I, poor response (0%–25% regression); Grade II, fair response (26%–50% regression); Grade III, good response (51%–75% regression); and Grade IV, excellent response (76%–100% regression). Grades III and IV were considered to have shown successful outcome. We graded the response similarly at the end of 6 months and in this series, there were one patient (1.9%) in Grade I, 7 (13.4%) in Grade II, 28 (53.8%) in Grade III, and 16 (30.7%) in Grade IV. Based on this criterion, the objective failure rate of response to propranolol therapy has been 15.38%.

Comparison between lesion site and response

Table 1 demonstrates the percentage reduction of the IHs at 1-, 2-, and 6-monthly intervals for the lesions at head and neck versus other sites. This analysis reveals that the location has a significant impact on clinical responses uniformly, favoring hemangiomas arising from the head and neck. After 6 months of therapy, the mean percentage reduction was 71.8 ± 12.6 for the head-and-neck group, while it was 51.7 ± 11.3 for the other sites (P ≤ 0.00001). Expectedly, the parental satisfaction was also higher for the head and neck group (P = 0.005) [Table 2].

Table 1

Mean percentage reduction of the infantile hemangiomas at 1-, 2-, and 6-monthly intervals for the lesions at head and neck versus other sites

	Site	Mean percentage reduction±SD	P
1 month	Head and neck	13.6±6.9	0.001
	Others	7.9±2.9
2 months	Head and neck	35.3±9.8	<0.00001
	Others	23.1±5.8
6 months	Head and neck	71.8±12.6	<0.00001
	Others	51.7±11.3

SD: Standard deviation

Table 2

Parental satisfaction levels at 6-month visit for lesions at head and neck versus other sites

Site	Parental satisfaction			Total
	Very satisfied (%)	Satisfied (%)	Unsatisfied (%)
Head and neck	26 (83.9)	4 (12.9)	1 (3.2)	31
Other sites	10 (47.6)	9 (42.8)	2 (9.5)	21
Total	36 (69.2)	13 (25)	3 (5.7)	52

Comparison between age at initiation of therapy and response

Analysis of age at treatment initiation and clinical response rates (for both cephalic and noncephalic sites) at the end of at least 6 months of therapy is shown in Table 3. It demonstrates that the percentage reduction in sizes is higher when the therapy is initiated before 2 years, even achieving statistical significance with P = 0.021 for the 0–6 months group and <0.0001 for 6 months–2 years group, respectively. In the cohort above 2 years of age, although the percentage reduction is greater at 36.25 ± 7.3 for the cephalic group, it fails to show statistical significance (P = 0.76).

Table 3

Analysis of age at treatment initiation and clinical response rates (for both cephalic and noncephalic sites) after 6 months

Site	Age group
	0-6 months	6 months-2 years	Above 2 years
Head and neck	79.64±10.5	66.25±9.9	36.25±7.3
Other sites	63.75±8.5	52.14±7.1	23.2±8.2
P	0.021	<0.0001	0.76

Comparison of responses between site of lesion/age at initiation and duration of therapy

The average treatment period in this study was 11.2 months (range, 2–16 months). It was observed that most lesions began to respond in size and coloration of the IHs within the 1st month of therapy. This effect was more pronounced in those who initiated therapy in the early proliferative phase. The lesions also continued to regress progressively after the rapid initial response. Although the duration was variable with response to the individual patient, the treatment was continued for at least 6 months, unless there were strong indication to discontinue prior (video above in the “Treatment regimen” section). It was noted that IHs located in the head and neck region required shorter duration of treatment (to achieve at least 50% regression in size) when compared with other locations (mean, 4.2 months and 8.5 months, respectively). Similar trend was noticed with duration and age at initiation (mean, 3.4 months for 0–6 month's age group and 7.3 months for >6 months age group).

Side/adverse effects

No severe adverse events were noted in our patients. Two (3.8%) patients had hypoglycemia that were easily managed. All other observed adverse effects were mild and were tolerated without discontinuing the medication [Table 4].

Table 4

Adverse drug reactions noted during the study period

Observed adverse effects	Numbers reported
Hypoglycemia	2
GI symptoms
Diarrhea	1
Vomiting	1
Constipation	0
Discomfort	1
Hypotension	0
Disturbances in sleep	0
Bronchospasm	1
Cardiac failure	0
Irritability	0
Seizure	0
Bradycardia	0
Cold extremities	0
Fatigue	0

GI: Gastrointestinal

DISCUSSION

After the serendipitous discovery of propranolol for the treatment of IHs,[8] a nonselective β-blocker used to treat hypertension, tachycardia, heart failure, and acute myocardial infarction, numerous reports around the world have described satisfactory responses.[91011] Before that systemic corticosteroids have been the mainstay of therapy for treatment of IHs.[12] Even though propranolol has become an attractive therapeutic alternative, there is no universally accepted protocol with propranolol or its dosing, and we have sparse literature about its safety and efficacy in the Indian population, especially infants. Through this work, we have endeavored to clarify certain aspects on its safety profile and end point of therapy, using a combination of subjective and objective outcome parameters to reduce biases. Although there are many studies on the outcomes with propranolol, very few of them actually compared its effect with respect to lesion site, age at initiation, or the duration of therapy.

The doses used in pediatric patients for this indication have ranged between 0.5 and 6 mg/kg/day.[13] In a meta-analysis by Marqueling et al. involving 41 studies of ≥1200 children treated with propranolol, with a mean dose of 2.1 mg/kg/day and treatment duration of 6.4 months, a response rate of 98% was noted. Serious side effects were rare, occurring in <1% of patients.[14] Consistent with these studies, we chose to use 2 mg/kg/day as the regular dose, with 1 mg/kg/day as the initiating dose in all age groups. About 3 mg/kg/day was used only if there has been nil response after at least 2 months of therapy. With this dosing range, propranolol has been safe and effective in our population.

The clinical and demographic characteristic of our patients were mostly in agreement with those previously described in the literature, except for the gender ratio of the affected patients which were roughly the same in our series. Based on our criteria for favorable clinical response, the success rate was determined to be 84.6% which correlates well with the parental satisfaction levels and the VAS scores. Although we had a mean duration of therapy as 11.2 months here, a consensus on the ideal duration of treatment required is not established. The authors have variably suggested to continue propranolol for 6 months duration, for at least the 1st year of life when the proliferation phase is mostly over, to complete resolution of the lesion.[15] In our experience, if the response to therapy has plateaued for 2 months after a treatment period of at least 6 months, regardless of the age at initiation, continuing it probably will not shrink it any further. But if the response is ongoing after 6 months, then the treatment should be continued till complete resolution or to the point of no further improvement. We have not had any rebounds or recurrences in the patients in this series after cessation of therapy thus far, but then some of them were on propranolol till recently. Tapering the dosage before cessation is perhaps one reason for this effect. Another reason could be that when the medication was started in infancy, we made sure not to withdraw it in the 1st year of life if there has been a response.

Location of the lesion was noted to be a significant factor affecting the therapeutic outcome in this study. Similar to some previous reports,[1617] the outcomes were better for the lesions in the head and neck region. The statistical analyses here revealed favorable responses that were uniformly higher at every stage (the mean percentage reduction at 6 months being 71.8 ± 12.6 for the head and neck group against 51.7 ± 11.3 for the other sites). The parental satisfaction and VAS were also noted to be favorable as would be expected. It has been suggested that propranolol exerts its effect as a nonselective β-adrenergic receptor (β2-AR) blocker that inhibits the cell growth and induces cell apoptosis of the endothelial cells.[18] The effect of location on the therapeutic outcome and treatment duration may be due to possible differences in β2-AR expression reportedly.[17] Perhaps, in this cohort of noncephalic group with suboptimal response, the second-line treatment options such as corticosteroids, vincristine, interferons, and surgical intervention should be considered.

Increased response rate in infants was also noted if the treatment began earlier than later. The group aged 0–6 months had higher resolution rates (percentage reduction wise) than the older groups. Besides, it was more evident in the cephalic subset. Further, we observed that the 0–6 months group required shorter treatment duration overall. For the cohort above 2 years, the percentage reduction in size was lower at 36.25 ± 7.3 in the cephalic lesions and 23.2 ± 8.2 in noncephalic lesions. This trend of better response in the proliferative phase when compared with involution phase (above 2 years) was also described previously.[1920] Zhang et al.[21] propose estrogen to be the critical factor that promotes angiogenesis through estrogen receptor α that in turn upregulates the expression of vascular endothelial growth factor A (VEGF-A) in the hemangioma-derived stem cell leading to proliferation of IHs. Propranolol inhibits this angiogenesis by downregulating the expression of VEGF. It is important therefore to not miss the window of opportunity during this critical period of early proliferation to intervene and prevent potential permanent sequelae, particularly in the cosmetically sensitive head-and-neck region. Especially when there are increasing reports of safety and efficacy of β-blockers and we become more comfortable with its use in very young infants.[171920] Notwithstanding this trend, propranolol is still useful as a first-line therapy at any age due to its long-term effect, namely, induction of apoptosis, that helps in regression of the IHs during the involution phase.[18]

The authors have used multiple and variable outcome measure to assess resolution of lesions. Since no objective laboratory value or other measures are established to determine size changes, investigators have developed multiple methods. A variety of scales, such as percentage change, mean change, VAS, have been used previously. Clearly, no single measure can be considered better than the other, though an objective parameter is desirable for comparisons. In this study, therefore, we have used a combination of subjective and objective outcome measures such as percentage reduction based on dimensions, VAS score, and parental satisfaction levels. Our success rate in this series has been 84.6% which is comparable with reported studies. A recent meta-analysis demonstrated an overall response rate of propranolol in IHs as 98% with a range of 82%–100%.[14] Similar efficacy has also been observed in a large prospective study by Hermans et al.[22] Price et al. determined a successful outcome as 81% in 59 patients (hemangioma clearance of 75% or more) over a mean duration of 7.9 months, by correlating percentage decrease in volume, cosmetically acceptable result by parent/physician, and a lack of need for further treatment.[23] Talaat et al. reported 75% clearance in 75% and >50% clearance in 94% of their 80 cases with IH in 5–8 months (mean, 6.53 ± 0.75).[3]

By close monitoring, after initiation of propranolol and after the first dose adjustment, our data demonstrated that the cardiac parameters were stable at the 1–3 mg/kg/day dosage. Although a reduction in the mean heart rate was noted, it was not sustained. All recorded blood pressures were within the normal range during therapy. Interestingly, results from recent case–control studies, including a randomized controlled trial of oral propranolol in 460 patients, showed that propranolol had no significant sustained effects on heart rates in infants with IHs.[24] These observations, together with our own observations, suggest that the cardiovascular risks of the use of propranolol in cutaneous IHs may be lower than initially feared. In the present study, any adverse event was recorded by parents between the study visits and was documented by investigators at each visit. But for a few side effects, most tolerated the drug at this dose with no single instance of discontinuing the medication due to an adverse event.

CONCLUSIONS

Propranolol has been shown to be safe and a promising therapeutic option in the treatment of cutaneous IHs. It has mild tolerable side effect with the dose used in this study and very few complications. The head-and-neck subset of IHs has shown better response in our series with shorter duration of therapy. Intervention during the early proliferative phase results in better outcome. However, randomized controlled studies are necessary to determine the optimal treatment dosage and duration, and recurrence rates, as well as to investigate any long-term adverse effects of propranolol.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.