Evidence-Based Medicine in the Treatment of Infantile Hemangiomas




Over the past decade, the treatment of infantile hemangiomas has undergone dramatic breakthroughs. This review critically evaluates the latest literature that supports the myriad treatment options for infantile hemangiomas. It chronicles the fading role of steroid therapy and evolution of propranolol use as the major treatment modality. Although propranolol is helping this disease become more of a medical disease and less of a surgical dilemma, the report also reveals a continued search to find nonsystemic treatment options. In summary, this is an evidence-based medicine review for the treatment of infantile hemangiomas.


Key points








  • The International Society for the Study of Vascular Anomalies (ISSVA) ( www.issva.org ) is the presiding organization for the classification of vascular anomalies.



  • Oral steroids are now considered a second choice option for the management of infantile hemangiomas (IHs).



  • Oral propranolol is becoming the first line of treatment of the management of IHs if observation or laser therapy is not sufficient.



  • Surgery has become more accepted as a means of early treatment of hemangiomas and relies on a physician’s clinical judgment regarding timing of intervention.



  • Multimodality algorithms, specifically addressing individual components of these tumors, result in the best functional and cosmetic outcomes.






Historical perspective and current classification scheme


The broader picture of vascular anomalies must first be elucidated to better understand the disease process of hemangiomas. The history of facial vascular anomalies is one plagued by confusing nomenclature, misdiagnosis, and lack of a unified consensus on the classification of an incredibly broad group of lesions. In 1982, however, the diligent work of Mulliken and Glowacki introduced the first classification system of vascular anomalies. Their study analyzed cellular characteristics of 49 vascular lesions and distinguished vascular tumors from vascular malformations based on histopathology, increased endothelial cell turnover, and differences in clinical history. A decade later, the ISSVA was founded and has since led the way in maintaining international consensus on the classification of vascular anomalies. The classification system has enhanced physician understanding and ability to accurately diagnose and appropriately prescribe therapy in patients afflicted with vascular anomalies.


The current classification scheme for vascular anomalies undergoes frequent updating and redefining by ISSVA ( Table 1 ). Currently, IH is classified as a benign vascular tumor. Like all vascular tumors/hemangiomas, IHs are characterized by increased endothelial cell turnover, which differentiates them from vascular malformations. Current theories surrounding the pathophysiology of IHs revolve around its unique expression of glucose transporter protein 1 (GLUT-1), possibly suggesting origin from placental endothelium, which is the only other cell in the body that expresses GLUT-1. The unique expression of GLUT-1 in IHs makes it an important marker for histologic diagnosis. IHs have been subclassified based on their diverse range of presentations. They are categorized based on depth of tissue involvement, including superficial, compound (mixed-type), and deep and reticular hemangiomas as well as by their patterns of distribution as either focal, multifocal, segmental, or indeterminate ( Table 2 ). Subclassification of these lesions is imperative to make appropriate treatment decisions.



Table 1

International Society for the Study of Vascular Anomalies classification for vascular anomalies




















Vascular Anomalies
Vascular Tumors Vascular Malformations
Simple Combined Of Major Named Vessels Associated with Other Anomalies



  • Benign vascular tumors




    • IH



    • Congenital hemangioma




      • Rapidly involuting congenital hemangioma a



      • Noninvoluting congenital hemangioma



      • Partially involuting congenital hemangioma




    • Tufted angioma



    • Spindle-cell hemangioma



    • Epithelioid hemangioma



    • Pyogenic granuloma (also known as lobular C hemangioma)



    • Others




  • Locally aggressive or borderline vascular tumors




    • Kaposiform hemangioendothelioma



    • Retiform hemangioendothelioma



    • Papillary intralymphatic angioendothelioma, Dabska tumor



    • Composite hemangioendothelioma



    • Kaposi sarcoma



    • Others




  • Malignant vascular tumors




    • Angiosarcoma



    • Epithelioid hemangioendothelioma



    • Others


C malformations
L malformations
V malformations
AV Ms a
AV fistula a
Defined as 2 or more vascular malformations identified in 1 lesion. Can be composed of any combination of: C, L, V, AV a Channel-type or truncal malformations Klippel-Trénaunay syndrome
Parkes Weber syndrome
Servelle-Martorell syndrome
Sturge-Weber syndrome
Limb CM + limb hypertrophy
Maffucci syndrome
Macrocephaly–CM
Microcephaly–CM
CLOVES syndrome
Proteus syndrome
Bannayan-Riley-Ruvalcaba syndrome

Abbreviations: AV, arteriovenous; C, capillary; CLOVES, congenital, lipomatous, overgrowth, vascular malformations, epidermal nevi, and spinal/skeletal anomalies and/or scoliosis; CM, capillary malformation; L, lymphatic; M, malformation; V, venous.

a High-flow lesions.



Table 2

Subclassification of infantile hemangiomas based on types and patterns of distribution





















Infantile Hemangioma
Pattern Types
Focal Superficial
Multifocal Mixed/compound
Segmental Deep
Indeterminate Reticular














Association with Other Lesions
PHACE association
Posterior fossa malformations, hemangioma, arterial anomalies, cardiovascular anomalies, eye anomalies, and sternal clefting and/or supraumbilical raphe
LUMBAR association
Lower body hemangioma, urogenital anomalies, ulceration, myelopathy, bony deformities, anorectal malformations, arterial anomalies, and renal anomalies


As evidence-based medicine has become the modern-day norm, it is now understood as the duty of the physicians to incorporate such evidence in their everyday practice and decision making. In treating facial vascular tumors, a host of therapeutic modalities exists, all with different levels of evidence supporting their efficacy, safety profiles, and utilization algorithms. This study attempts to shed light on the state of the current evidence supporting different treatment modalities for IHs.




Historical perspective and current classification scheme


The broader picture of vascular anomalies must first be elucidated to better understand the disease process of hemangiomas. The history of facial vascular anomalies is one plagued by confusing nomenclature, misdiagnosis, and lack of a unified consensus on the classification of an incredibly broad group of lesions. In 1982, however, the diligent work of Mulliken and Glowacki introduced the first classification system of vascular anomalies. Their study analyzed cellular characteristics of 49 vascular lesions and distinguished vascular tumors from vascular malformations based on histopathology, increased endothelial cell turnover, and differences in clinical history. A decade later, the ISSVA was founded and has since led the way in maintaining international consensus on the classification of vascular anomalies. The classification system has enhanced physician understanding and ability to accurately diagnose and appropriately prescribe therapy in patients afflicted with vascular anomalies.


The current classification scheme for vascular anomalies undergoes frequent updating and redefining by ISSVA ( Table 1 ). Currently, IH is classified as a benign vascular tumor. Like all vascular tumors/hemangiomas, IHs are characterized by increased endothelial cell turnover, which differentiates them from vascular malformations. Current theories surrounding the pathophysiology of IHs revolve around its unique expression of glucose transporter protein 1 (GLUT-1), possibly suggesting origin from placental endothelium, which is the only other cell in the body that expresses GLUT-1. The unique expression of GLUT-1 in IHs makes it an important marker for histologic diagnosis. IHs have been subclassified based on their diverse range of presentations. They are categorized based on depth of tissue involvement, including superficial, compound (mixed-type), and deep and reticular hemangiomas as well as by their patterns of distribution as either focal, multifocal, segmental, or indeterminate ( Table 2 ). Subclassification of these lesions is imperative to make appropriate treatment decisions.



Table 1

International Society for the Study of Vascular Anomalies classification for vascular anomalies




















Vascular Anomalies
Vascular Tumors Vascular Malformations
Simple Combined Of Major Named Vessels Associated with Other Anomalies



  • Benign vascular tumors




    • IH



    • Congenital hemangioma




      • Rapidly involuting congenital hemangioma a



      • Noninvoluting congenital hemangioma



      • Partially involuting congenital hemangioma




    • Tufted angioma



    • Spindle-cell hemangioma



    • Epithelioid hemangioma



    • Pyogenic granuloma (also known as lobular C hemangioma)



    • Others




  • Locally aggressive or borderline vascular tumors




    • Kaposiform hemangioendothelioma



    • Retiform hemangioendothelioma



    • Papillary intralymphatic angioendothelioma, Dabska tumor



    • Composite hemangioendothelioma



    • Kaposi sarcoma



    • Others




  • Malignant vascular tumors




    • Angiosarcoma



    • Epithelioid hemangioendothelioma



    • Others


C malformations
L malformations
V malformations
AV Ms a
AV fistula a
Defined as 2 or more vascular malformations identified in 1 lesion. Can be composed of any combination of: C, L, V, AV a Channel-type or truncal malformations Klippel-Trénaunay syndrome
Parkes Weber syndrome
Servelle-Martorell syndrome
Sturge-Weber syndrome
Limb CM + limb hypertrophy
Maffucci syndrome
Macrocephaly–CM
Microcephaly–CM
CLOVES syndrome
Proteus syndrome
Bannayan-Riley-Ruvalcaba syndrome

Abbreviations: AV, arteriovenous; C, capillary; CLOVES, congenital, lipomatous, overgrowth, vascular malformations, epidermal nevi, and spinal/skeletal anomalies and/or scoliosis; CM, capillary malformation; L, lymphatic; M, malformation; V, venous.

a High-flow lesions.



Table 2

Subclassification of infantile hemangiomas based on types and patterns of distribution





















Infantile Hemangioma
Pattern Types
Focal Superficial
Multifocal Mixed/compound
Segmental Deep
Indeterminate Reticular














Association with Other Lesions
PHACE association
Posterior fossa malformations, hemangioma, arterial anomalies, cardiovascular anomalies, eye anomalies, and sternal clefting and/or supraumbilical raphe
LUMBAR association
Lower body hemangioma, urogenital anomalies, ulceration, myelopathy, bony deformities, anorectal malformations, arterial anomalies, and renal anomalies


As evidence-based medicine has become the modern-day norm, it is now understood as the duty of the physicians to incorporate such evidence in their everyday practice and decision making. In treating facial vascular tumors, a host of therapeutic modalities exists, all with different levels of evidence supporting their efficacy, safety profiles, and utilization algorithms. This study attempts to shed light on the state of the current evidence supporting different treatment modalities for IHs.




Epidemiology and clinical presentation


IHs are the most common benign vascular tumors of infancy and childhood, occurring in 4.5% of all infants by age 3 months and in up to 12% by 1 year of age. IHs occur between 3 and 6 times more frequently in girls than boys; are seen more often in whites, premature infants, and twins; and tend to afflict offspring of mothers of higher maternal age. Approximately 60% of these occur in the head and neck region, 25% on the trunk and 15% on the extremities. At birth, most IHs are absent or have a small premonitory mark, such as a pale macule. By 1 to 4 weeks of age, most IHs have become clinically apparent. Depending on depth of involvement, IHs have a diverse appearance; superficial IHs tend to be bright red, slightly elevated, noncompressible plaques, whereas deep IHs are soft, warm, and bluish in color. Compound hemangiomas develop both deep and superficial components ( Fig. 1 ). Focal and multifocal lesions present in 1 or multiple locations, respectively, whereas segmental hemangiomas pattern within facial dermatomes. Segmental hemangiomas are classically associated with PHACE (association of posterior fossa brain malformations, hemangiomas, arterial anomalies, coarctation of the aorta and cardiac defects, and eye abnormalities) syndrome in anywhere from 20% to 31% of cases.




Fig. 1


Different types and patterns of distribution of IHs. ( A ) Superficial hemangioma of the postauricular skin, ( B ) compound hemangioma of the left eyebrow, ( C ) deep hemangioma of the upper lip, and ( D ) segmental hemangioma of the face warranting work-up up of PHACE diagnosis.


PHACE syndrome is characterized by the presence of a segmental or large facial hemangioma associated with posterior fossa and cerebral vascular anomalies as well as cardiac and ocular abnormalities. A consensus group meeting in 2009 defined diagnostic criteria for PHACE syndrome. These infants require additional work-up, including further imaging of the head, neck, and chest and dermatologic and ophthalmologic examinations. Acute ischemic stroke is a known risk in these individuals given their predilection for underlying vascular anomalies and must be considered when waging optimal treatment regimens for their tumors.


Clinical course of IHs varies widely but typically follows a consistent pattern of proliferation, latency, and involution. After initiation of the proliferative phase of growth, IHs have been shown to reach approximately 80% of their final size by a mean age of 3 to 6 months. By 12 months, a vast majority of IHs have entered the latent phase or have already began to involute, although there are rare cases where these tumors continue to grow past 2 years of age. As the involution phase takes over, IHs drastically shrink in size and turgidity and experience color change from red to gray and often return to a more neutral skin color over the next 5 to 7 years.


A majority of IHs follow the described natural history of disease without need for intervention and have mostly involuted by late childhood (approximately 50% by age 5% and 70% by age 7). Approximately 40% to 50% of children, however, are left with either residual tumor, loose and irregular skin, telangiectasias, or scarring, which may require intervention down the road for cosmesis. It is estimated that at some point during their clinical course, approximately 10% of IHs become destructive or disfiguring or may threaten vision or airway patency; become ulcerated or infected; cause heart failure, severe thrombocytopenia, or bleeding; and, thus, be life-threatening. These clinical scenarios often warrant intervention. There have been several studies that have looked at the association of clinical characteristics and predilection for developing complications to better inform need for treatment and specialty referral—among these worrisome features are facial location, segmental morphology, and larger size (for every 10 cm 2 of increasing size, there is a 5% increase in the likelihood of developing a complication).


Given the great diversity in presentation and clinical course of these lesions, no single treatment option is ideal for all cases. The overarching goal of intervention for IHs is to maximize the functional and cosmetic result while minimizing associated morbidity and complications. Ideally this treatment goal has been achieved by 3 to 5 years of age when children have developed self-image, experience new social pressures associated with school, and have become more psychologically vulnerable. Evidence-based therapies for the treatment of IHs re reviewed in detail later.




Watch and wait


As discussed previously, a majority of IH patients (approximately 60%) do not require medical or surgical treatment because their lesions are self-involuting and never pose permanent cosmetic, functional, or psychological morbidity. In these cases, a conservative approach of parental reassurance, close follow-up, and routine monitoring for clinical changes is sufficient. In general, it is acceptable practice to watch these lesions until intervention is predicted to lead to a better outcome than allowing a tumor to follow its natural course of disease. In cases where surgical intervention might be warranted, waiting for complete involution may offer the benefit of increased skin availability, better scar location and size, and smaller risk of damage to adjacent structures. With the advent of propranolol, the modern-day trend is a gravitation of parents and practitioners toward early pharmacologic intervention for a greater range of tumors than ever before, and, therefore, the watch-and-wait approach may be losing popularity.




Steroids—systemic/intralesional


Until the first use of propranolol in 2008, both systemic and IL steroids had been a mainstay of treatment of IHs. Systemic and intralesional (IL) steroids had proved effective but not without causing substantial side effects, although mainly short term, including growth disturbances, infections, personality changes, cushingoid facies, and immunosuppression. The majority of literature supporting steroid use is level IV–V evidence in the form of small case series and observational and retrospective studies. A study by Enjolras and colleagues showed that steroids bring about significant tumor improvement in approximately 30% of patients, another 30% show no change and ultimately require dose increases to 5 mg/kg per day, and 40% experience partial/minimal responses (level V evidence). Pandey and colleagues reviewed their institution’s 20 year experience with these modalities (level IV evidence). A total of 2013 patients received either oral steroids (499 patients), IL steroids (886 patients), or a combination for these modalities (628 patients). Results revealed that superficial hemangiomas responded best to all treatment modalities—95% had greater than 75 to 50% regression versus 90% of mixed and 70% of deep lesions ( P <.01 for superficial vs deep responses). Children less than 1 year of age showed statistically significant better response than older children. The complications of treatment were local infections related to ulceration (12.4%, not statistically different between the 3 groups); cushingoid facies/growth delay and hypertension (3.1% and 2.5%, respectively, statistically more frequent in the dual-modality and systemic steroid groups). The investigators recommended that for smaller lesions (≤25 cm 2 ), IL therapy should be first choice, regardless of hemangioma type (superficial, deep, or mixed). Oral therapy was recommended as the initial choice for larger (≥25 cm 2 ) or multiple tumors and combined therapy for nonresponders to IL therapy. Many other investigators have provided level V evidence that is in agreement with Pandey and colleagues’ advocacy for the use of IL steroids for small localized hemangiomas where directed injections can cause tumor response without the systemic side effects of other pharmacotherapies.


Recognizing the lack of higher-level evidence supporting steroid use for IHs, Bennett and colleagues designed a meta-analysis in 2001 compiling data from 10 case series, including 184 patients with IHs who were treated with systemic glucocorticoids (level III evidence). The investigation found a response rate of 84% and a rebound rate of 36%. Adverse events were reported in 34% of cases independent of dosing. In the first small RCT to provide level II evidence looking at efficacy of oral steroids for IH treatment, Pope and colleagues randomized 20 patients with problematic IHs to either daily oral prednisolone or monthly intravenous pulses of methylprednisolone. Improvement using a visual analog score that estimates overall changes in size and adverse events were compared at 3 months from baseline and 1 year of age. The investigators found that oral prednisone therapy resulted in improved resolution but had a higher rate of adverse events compared with pulse dosing.


Highlighting the substantial difference in safety profile between corticosteroids and propranolol, Sawa and colleagues performed a prospective database analysis (level IV evidence) to describe the long-term effects of steroids versus β-blockers on anthropometric measurements in 18 children with IHs. The investigators found a significant increase in body mass index and decrease in height percentile in children treated with steroids. The propranolol group required longer duration of therapy and had significantly lower systolic blood pressures but without symptomatic hypotension. Additionally, Price and colleagues supported propranolol’s efficacy to be superior to steroids and to lead to significantly less relapses and less need for surgical intervention as well as lower cost and side effects (level IV evidence). Although substantial evidence exists supporting the efficacy of steroids for management of IHs, their poor side-effect profile is well-supported in the literature and has led to a dramatic shift in management over the past 5 years toward β-blocker therapy as the preferred treatment modality for IHs. This shift is not only obvious when reviewing the literature but also has been demonstrated to exist within individual institutions as is shown by level V evidence from Gomulka and colleagues.




Propranolol


Propranolol hydrochloride, a nonselective β-blocker, is rapidly becoming the mainstay of treatment of IHs. In a landmark New England Journal of Medicine article from 2008, Léauté-Labrèze and colleagues published a small case series inspired by patients treated with propranolol for cardiopulmonary purposes who were incidentally noted to have response of their IHs to the drug (level V evidence). Within a year of this first report by Léauté-Labrèze and colleagues, several level V reports further confirmed the efficacy of propranolol for management of airway IHs. A subsequent series of 32 cases by Sans and colleagues reported immediate effect of propranolol treatment on color and growth of IHs leading to considerable shortening of the course of disease (level V evidence). Whereas most prior studies had looked at treatment of IHs in the proliferative phase, Zvulunov and colleagues’ retrospective study showed that latent or involuting tumors treated with propranolol experience faster rates of involution compared with untreated tumors (level V evidence). The first larger 2010 retrospective review by Buckmiller and colleagues showed that among 32 patients with IHs, 97% demonstrated response to propranolol therapy and half did not require any further treatment (level V evidence). These studies reported that therapy was overall well tolerated, with only minor side effects, such as gastrointestinal reflux disease, somnolence, respiratory syncytial virus exacerbation, and rash. Several investigators continued to report their experiences in 2010 to 2011 through mainly level V evidence, including case reports and series, all supporting the efficacy and excellent safety profile of propranolol and endorsing its use as a first-line therapy for management of these tumors ( Fig. 2 ). As the efficacy and safety profile of propranolol became largely supported by level IV–V evidence during the early years after its discovery in 2008, higher levels of evidence were still needed to justify its widespread use.




Fig. 2


Example of a left ear IH treated with propranolol. ( A ) Left ear appearance at 3 months of age prior to propranolol treatment. ( B ) Left ear appearance after 3 months of propranolol treatment.


After these observational and retrospective studies, several investigators recognized the need for a large meta-analysis and randomized controlled trials (RCTs) to further prove efficacy, and establish criteria for administration. Hogeling and colleagues compared 24 weeks of propranolol with placebo in the treatment of 40 children ages 9 weeks to 5 years with facial or potentially disfiguring IHS (level II evidence). In this RCT, the investigators found a significant reduction in IH redness and elevations at weeks 12 and 24 of treatment in the propranolol group versus placebo ( P <.01 and P <.001, respectively), with no serious adverse effects reported.


A meta-analysis from 2011 compiling 213 patients from 49 articles described a common approach and course of treatment when using propranolol for management of IHs (level III evidence). In this study, 93% of patients had therapy initiated in infancy at a mean age of 4.5 months. Initiation of therapy was closely monitored in all studies but with different approaches—some practiced inpatient monitoring whereas others implemented close outpatient monitoring with frequent follow-up. Dosing in 65% of patients was 2 mg/kg/d, and in 25% of patients 3 mg/kg/d, for an average duration of 5.1 months. Without any clear guidelines for propranolol use at the time, this meta-analysis unveiled the discrepancies in how treating physicians were managing IHs with β-blockers and underscored the need for more uniform consensus.


There have been several more recent, well-designed meta-analyses and RCTs conducted in the past few years showing direct superiority of propranolol over steroids for IHs. Malik and colleagues designed a RCT comparing the efficacy of orally administered propranolol versus systemic steroids versus using both medications concurrently. The investigators showed that propranolol brought about earlier improved changes and that there was no significant increased benefit of adding prednisolone to propranolol, but this addition did lead to poorer patient compliance secondary to a higher number of complications associated with steroid use. A meta-analysis in 2013 pooled data from 1965 to 2012, resulting in a comparison of 2697 patients from 16 studies who had received oral or locally administered glucocorticoids with 795 pooled patients from 25 studies who had been treated with propranolol (level II evidence). Systemic glucocorticoids had an overall efficacy of 71% versus 97% for treatment with propranolol. Furthermore, the study revealed that steroid complication rate was double that of propranolol therapy (23% vs 9.6%, respectively). Most recently, a 2014 phase 2, investigator-blinded, multi-institutional RCT was published comparing propranolol and prednisolone in the management of IHs. The study was terminated prior to targeted enrollment secondary to adverse events, leading to withdrawal in the prednisolone-treated group (level II evidence).


Overall, there exists substantial level IV–V evidence in the form of observational and retrospective studies as well as some level II–III evidence in the form of meta-analyses and RCTs, providing physicians with adequate evidence-based support for the efficacy and safety of propranolol as a first-line treatment of IHs. In return, the Food and Drug Administration (FDA) has recently approved Hemangeol, (Pierre Fabre Pharmaceuticals, Inc, Parsippany, NJ, USA) a liquid suspension of propranolol, as the first-ever FDA approved treatment option for proliferating IHs. This approval was based largely on double-blinded RCT. In this phase II/III trial, 460 infants were randomized to receive placebo or 1 of 4 dosing regimens (1.2 or 3.4 mg/kg/d in twice-daily divided doses for 3 or for 6 months) of Hemangeol. At 24 weeks, treatment response was evaluated by blinded, centralized, independent assessments of photographs compared with baseline. The trial showed that 60% of infants treated with Hemangeol versus 4% of those treated with placebo, met the primary endpoint of complete or near-complete resolution of the IH ( P = .0001). The most frequently reported adverse reactions in the treatment group were sleep disorders (16.1% vs 5.9% in placebo) and aggravated respiratory tract infections, such as bronchitis and bronchiolitis (13% vs 4.7% in placebo), leading to treatment discontinuation in fewer than 2% of patients. Current absolute contraindications for the use of Hemangeol are prematurity (corrected age <5 weeks), infants weighing less than 2 kg, known hypersensitivity to propranolol, asthma or history of bronchospasm, heart rate less than 80 beats per minute, decompensated heart failure, greater than first-degree heart block, blood pressure less than 50/30 mm Hg, and pheochromocytoma ( Box 1 ). A barrier to its widespread clinical use is its exceedingly high cost, which raises the question of whether Hemangeol has any increased efficacy over generic liquid propranolol and other β-blocker alternatives. In a recent, small RCT, 23 IH patients were randomized to receive either propranolol or atenolol and a noninferiority analysis was conducted. When comparing complete responses (60% in propranolol group and 53.8% in the atenolol group; P = .68) and side-effect profiles of these 2 β-blockers in the treatment of IHs, there were no significant differences.


Feb 8, 2017 | Posted by in General Surgery | Comments Off on Evidence-Based Medicine in the Treatment of Infantile Hemangiomas

Full access? Get Clinical Tree

Get Clinical Tree app for offline access