Objectives: To systematically review evidence addressing the diagnosis and management of infantile hemangiomas (IH).
Data sources: Multiple databases from 1982 to June 2015.
Review methods: We included comparative studies of interventions, case series addressing harms, and any study to address contextual questions. Two investigators independently screened studies and rated study quality. We extracted and summarized data qualitatively and quantitatively via network meta-analysis, which provides a relative ranking of anticipated effects among treatments. We also assessed strength of the evidence (SOE).
Results: Among 148 unique studies, 42 addressed effectiveness outcomes (6 good, 22 fair, and 14 poor quality), and 144 studies reported harms (14 good, 3 fair, and 127 poor quality). Two small studies reported differing findings for the sensitivity of ultrasound and effectiveness of imaging modalities. Studies of steroids assessed different agents; treated children typically had improvement in lesion size. Steroid harms frequently included Cushingoid facies, irritability/mood changes, and growth retardation. Beta-blockers typically demonstrated significantly greater effects on reducing lesion size than did control or other active comparators. In network meta-analysis, oral propranolol had the largest mean estimate of expected clearance (95%; 95% Bayesian credible interval [BCI]: 88% to 99%) relative to oral corticosteroids (43%, 95% BCI: 21%–66%) and control (6%, 95% BCI: 1%–11%). Beta-blocker harms included hypotension, hypoglycemia, bradycardia, sleep disturbances, and cold extremities. Surgical intervention studies primarily addressed variations of pulse dye laser (PDL) to manage IH size. Most studies reported a higher success rate with longer-pulse PDL compared to observation, with differing magnitude of effect. Laser treatment harms included hypopigmentation and scarring. No studies explicitly evaluated treatments following failure of beta-blockers or corticosteroids. Literature addressing contextual questions suggested that referral indications include large size; segmental type; risk for complications including bleeding, ulceration, and pain; involvement of critical structures; risk factors for occult lesions (numerous cutaneous lesions, beard distribution); and potential for psychosocial concerns in some cases. Multiple case series reported associations between multiple cutaneous lesions and airway or hepatic IH and facial lesions in a beard distribution and airway IH.
Conclusions: Our review for contextual questions described a range of indications for referral and suggested support for a higher index of suspicion of extracutaneous IH in children with multiple cutaneous lesions or facial lesions in a beard distribution. Corticosteroids demonstrated moderate effectiveness at reducing IH size/volume (moderate SOE for improvement in IH with oral steroids compared with observation/placebo; low SOE for intralesional steroids versus observation/placebo; moderate SOE for association with clinically important harms). Propranolol had high SOE for effects on reducing lesion size compared with observation/placebo. Clearance of IH was greater in propranolol arms compared with placebo/observation and active comparators in most studies. Meta-analysis indicated high mean rates of IH clearance with oral propranolol (95%, 95% BCI: 88%–99%) and moderate rates for steroids (43% to 58%, with wide BCI; moderate SOE for effects of propranolol compared with steroids). Beta-blockers and steroids also may cause clinically important harms (moderate SOE for association of oral propranolol with harms). Laser studies generally found PDL more effective than other lasers, but effects remain unclear (insufficient to low SOE for effects of laser types on IH clearance; moderate SOE for association of PDL with skin pigmentation changes; low SOE for association with pain). Data were inadequate to address the role of imaging in guiding treatment (insufficient SOE).