TOPLINE

No active intervention demonstrated superiority over oral propranolol for treating infantile hemangioma (IH). Corticosteroids were associated with substantially higher adverse event rates compared with propranolol.

METHODOLOGY

• Researchers conducted a Bayesian network meta-analysis of 30 randomized controlled trials published from January 2008 to June 2026.
• A total of 2639 patients across 9 treatment nodes were included, and 1143 patients across 8 nodes were included in the safety analysis from 12 RCTs.
• Treatment included oral propranolol, atenolol, nadolol, intralesional propranolol, topical beta-blockers, laser therapy, corticosteroids, propranolol combination therapy, and placebo/observation.
• The primary efficacy outcome was treatment success rate, defined as complete lesion resolution or at least 75% reduction in size at the assessment endpoint, typically at 6 months post-initiation.

TAKEAWAY

• No active treatment demonstrated superiority over propranolol, whereas placebo/observation was significantly inferior (odds ratio [OR], 0.12; 95% credible interval [CrI], 0.03-0.52).
• Corticosteroids were associated with higher rates of adverse event compared with propranolol (OR, 52.92; 95% CrI, 3.12-2874.40), while placebo showed significantly lower adverse event odds (OR, 0.14; 95% CrI, 0.03-0.57).
• Substantial between-study heterogeneity was observed in the efficacy network.

IN PRACTICE

"By anchoring all comparisons to oral propranolol and integrating CINeMA certainty assessments with inconsistency testing, these findings are consistent with the continued role of oral propranolol as the reference systemic treatment for infantile hemangioma. For patients intolerant to propranolol, atenolol may represent a reasonable alternative," wrote the authors of the study.

SOURCE

The study was led by Caihong Li, University Hospital Erlangen, Friedrich-Alexander University of Erlangen-Nürnberg, Erlangen, Germany. It was published online on July 21 in European Journal of Pediatrics.

LIMITATIONS

Substantial heterogeneity and global inconsistency were observed in the efficacy network. Several treatment nodes were supported by limited trials with wide credible intervals. More than half of included trials carried high risk of bias.

DISCLOSURES

The study was funded through Projekt DEAL. The authors reported having no relevant conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.