Complications of systemic corticosteroid therapy for problematic hemangioma.

Abstract

Systemic corticosteroid therapy has been used to treat hemangiomas for 30 years; yet, there are no studies of possible complications. We reviewed the database of the Vascular Anomalies Center at the Boston Children's Hospital and gathered information on short- and long-term side effects in children who were given systemic corticosteroids for problematic hemangiomas. In addition, a questionnaire regarding early and late consequences was sent to the families of children who were treated with corticosteroids from 1983 to 1997. Of 300 patients with hemangiomas, 80 children were identified as having received a full course of systemic corticosteroids for problematic tumors. Complete data were collected on 62 of these children. The response rate to the questionnaire was 78 percent (n = 62 of 80). The initial dose of corticosteroid varied from 2 to 3 mg/kg/ day. Duration of therapy ranged from 2 to 21 months (mean, 7.9 months; median, 6.5 months). The follow-up interval from the cessation of therapy ranged from 6 months to 15 years (mean, 4 years; median, 3 years). Short-term complications included cushingoid facies (n = 44; 71 percent), personality changes (n = 18; 29 percent), gastric irritation (n = 13; 21 percent), fungal (oral or perineal) infection (n = 4; 6 percent), and diminished gain of height (n = 22; 35 percent) and weight (n = 26; 42 percent). A total of 91 percent of children who had diminished gain of height (n = 20) returned to their pretreatment growth curve for height by 24 months of age. One child, who was treated at another institution with a dose of 20 mg/kg/day for 6.5 months that was slowly tapered over 18 months, was petite 6 years after ending therapy. Another child treated with an initial dose of 2 mg/kg/day for 5 months was smaller than predicted at the age of 6 years, but she was born prematurely and was on ventilatory support for respiratory distress. Three children treated with the standard dose and duration were at a low percentile for weight 4, 5, and 10 years after the cessation of therapy. Statistical analysis showed a correlation between diminished gain of height with duration of therapy and age at initiation of treatment. One child had corticosteroid myopathy that resolved with cessation of therapy. We found no evidence for immunologic suppression, i.e., there was no increase in the number of bacterial infections during corticosteroid administration. In conclusion, systemic corticosteroids can be safely given to treat endangering hemangiomas in infants at doses of 2 to 3 mg/kg/day, which are slowly tapered and stopped before the age of 1 year. Short-term side effects were minor and transient, and no serious long-term complications occurred.