Infantile haemangioma

Meet the patient

A 6-week-old girl is brought in by anxious parents because a "red spot" on her cheek, barely visible at birth, has mushroomed into a bright red, raised, lobulated strawberry plaque the size of a five-cent coin in three weeks. The parents have been told it is a birthmark that will not change; they can see it changing weekly. They want it removed.[1][6]

The two questions that settle their next five years: is this going to keep growing, and will it go away? (yes, then yes — the natural history of IH) and is this a high-risk lesion that needs propranolol now, or a low-risk one I watch? Most are low-risk; the skill is spotting the minority that are not.[1][3]

Tumour, not malformation — the ISSVA fork

The first classification decision is the one that prevents years of wrong management: is this a vascular tumour or a vascular malformation? They look alike to the untrained eye; they behave nothing alike.[5]

[1] [5]

The single immunostain that settles it when the clinical picture is ambiguous is GLUT1 (glucose transporter-1): positive in infantile haemangioma, negative in congenital haemangiomas and vascular malformations. Likely of placental origin (the placental hypoxia hypothesis), IH is driven during proliferation by VEGF, FGF-2, beta-catenin and IGF-2 from a CD133-positive stem-cell population, and replaced by fibrofatty tissue during involution.[4][5]

Proliferate, plateau, involute — the defining feature

Memorise the three phases, because the natural history is the diagnosis and the treatment plan in one. IH is absent or a precursor macule at birth, then:[1][6]

1. Proliferative phase (0–5 months) — rapid growth; about 80 percent of the final size is reached by 5 months, most of it in the first three.
2. Plateau phase (around 6–12 months) — growth stops; the lesion stabilises.
3. Involution phase (1–9 years) — slow spontaneous regression; roughly 30 percent resolved by age 3, 50 percent by 5, 70 percent by 7, 90 percent by 9. Residual telangiectasia, atrophy or fibrofatty tissue may remain in 20–40 percent.[1]

The classic trap — "it grew, so it must be a malformation": vascular malformations are present and fully formed at birth and grow only proportionately with the child. A lesion that was absent at birth and is proliferating postnatally is an IH until proven otherwise. The inverse trap — assuming a fully-grown-at-birth lesion is an IH — points to a congenital haemangioma (RICH rapidly involutes by 12–18 months; NICH never involutes; PICH partially), which is GLUT1-negative and does not respond to propranolol.[4][5]

How propranolol works — the three-phase mechanism

The drug that transformed this field works in three sequential waves, and a fellowship candidate must name each.[3][4]

1. Early (hours to days) — vasoconstriction: beta-2 blockade cuts endothelial nitric oxide, so the lesion visibly flattens and softens within 24–48 hours. This is the effect parents see first.
2. Intermediate (weeks) — anti-angiogenesis: blockade of VEGF and FGF signalling halts endothelial proliferation and stops the growth.
3. Late (months) — apoptosis: beta-1 blockade disables the Akt/PI3K survival pathway, driving caspase-mediated death of capillary endothelial cells and the long-term regression.[3]

Recognise the high-risk lesion

Most IH are uncomplicated; the skill is identifying the minority that threaten function or looks — and treating them early, in the proliferative phase when propranolol works best. About 10–20 percent of IH develop complications; about 60 percent sit on the head and neck.[1][3]

[1] [11]

The ulceration warning sign everyone forgets: an IH that develops a pale white discoloration is signalling impending ulceration — escalate to propranolol and protective care before the skin breaks, not after. Ulceration is most common on the lip, perineum and neck, and on large, superficial, segmental and lower-lip lesions.[11][12]

PHACES and LUMBAR — when the skin lesion is the tip

A large segmental IH is not just a skin lesion; it is the visible clue to a syndrome, and missing the screen is the recurring failure.[1][2]

[1] [2]

A lumbosacral segmental IH carries the mirror syndrome — LUMBAR (Lower body IH, Urogenital, Myelopathy, Bony, Anorectal, Renal anomalies) — and needs spinal MRI for a tethered cord.[1]

The Kasabach-Merritt trap

Kasabach-Merritt phenomenon — thrombocytopaenia with a consumption coagulopathy in an enlarging vascular lesion — is caused by kaposiform haemangioendothelioma or tufted angioma, NOT by infantile haemangioma. This is the classic exam trap, and getting it wrong sends a child down the wrong pathway. IH does not cause Kasabach-Merritt; sirolimus and vincristine, which treat KHE, are not drugs for ordinary IH.[4][5]

Differential diagnosis — the face-off

Four mimics account for most confusion; the discriminator is behaviour over time plus GLUT1.[1][5]

[1]

Investigations — mostly clinical

IH is a clinical diagnosis from the history (absent at birth, proliferating) and morphology — most need no test at all. Imaging is reserved for deep, large or ambiguous lesions and for syndrome screening.[1][13]

• Ultrasound with Doppler — for deep or subcutaneous lesions, and to separate IH from a malformation (a high-flow lesion).
• MRI — for large, deep or ambiguous lesions, and mandatory for PHACES (brain and MRA) or LUMBAR (spine).
• GLUT1 immunostaining on biopsy — confirms IH when the diagnosis is genuinely uncertain.
• Abdominal ultrasound when there are 5 or more cutaneous IH, to screen the liver for hepatic haemangiomatosis.
• FBC and coagulation only if Kasabach-Merritt is suspected (and remember it is KHE, not IH).[3][13]

Management — the ladder, risk-stratified

Treat the high-risk lesion early in the proliferative phase; leave the low-risk one alone and reassure the parents. The AAP 2019 Clinical Practice Guideline set the modern standard, and it is explicit: propranolol first-line for high-risk, topical timolol for small superficial, steroids deprecated.[3][15]

[3]

Active non-intervention (most IH)

Reassure the parents with the numbers — about 50 percent resolved by age 5, 90 percent by 9 — photograph for monitoring, and review every 1–3 months through the proliferative phase. Treat residual telangiectasia or fibrofatty tissue later if it troubles the child.[1][6]

Oral propranolol — first-line systemic for high-risk IH

The target is 2–3 mg/kg/day divided twice daily, continued through the proliferative phase to about 12–15 months of age. The AAP 2019 titration: start at 0.5–1 mg/kg/day, increase by 0.5 mg/kg every 1–2 days to target, monitoring heart rate and blood pressure. A pretreatment ECG is a strong recommendation; echocardiogram if the history or exam suggests cardiac disease; and initiate the first dose in a clinical setting for infants under 3 months, with cardiac disease, or with airway involvement.[3][6][15]

The two monitoring traps. First, hypoglycaemia — propranolol masks the adrenergic warning signs, so hold the dose if the infant is vomiting or fasting (any intercurrent illness), and give doses with feeds. Second, PHACES infants titrate more slowly — they carry a stroke risk from hypotension, so image the brain and arteries first and ramp gently. Contraindications are severe asthma or bronchospasm, heart block, severe heart failure, and hypoglycaemia risk.[3][9][14]

[1] [15]

Topical timolol, atenolol, and the alternatives

Topical timolol 0.5 percent gel-forming solution, one drop twice daily on the lesion, is for small (under 1 cm), thin, superficial IH where systemic therapy is unnecessary; systemic absorption is negligible, with 47–88 percent improvement in published series. Avoid it on ulcerated IH (rapid absorption through denuded skin) and in preterm infants.[3][10][14]

Atenolol — a beta-1-selective blocker, 0.5 mg/kg/day titrated to 1 mg/kg/day once daily — is non-inferior to propranolol (the Ábarzúa-Araya 2014 RCT) with fewer sleep disturbances and less bronchospasm, making it the choice for propranolol-intolerant infants or those with reactive airways; it remains off-label.[16]

Captopril is an emerging third-line for the infant who cannot take any beta-blocker (severe asthma, bradycardia, prior propranolol failure): start 0.1 mg/kg every 12 hours and up-titrate to 2 mg/kg every 12 hours, monitoring blood pressure, renal function and potassium. It is not yet first-line in any guideline.[17]

Systemic corticosteroids (prednisolone 2–3 mg/kg/day) were the pre-2008 first-line and are now second-line for propranolol failure or contraindication, with growth retardation, hypertension and immunosuppression as the price. The AAP 2019 explicitly deprecates them for uncomplicated IH. Surgery and pulsed dye laser are reserved for residual skin change after involution, small pedunculated lesions, and obstruction unresponsive to medical therapy.[3][15]

The trials that changed practice

The propranolol era began by serendipity in Bordeaux in 2007: two infants given propranolol for hypertrophic cardiomyopathy coinciding with severe facial IH watched their haemangiomas soften and flatten within 24 hours. The 2008 NEJM letter extended this to 11 more infants, all improved, none needing steroids — a chance observation that rewrote a specialty.[8]

The evidence that made it standard of care was the Léauté-Labrèze 2015 NEJM trial — 460 infants randomised to placebo, propranolol 1 mg/kg/day, or propranolol 3 mg/kg/day for 6 months. At the 3 mg/kg/day dose, 60 percent had complete or near-complete resolution versus 4 percent on placebo; 96 percent showed some improvement; serious adverse events were 1 percent (a single hypoglycaemia). That dose became the FDA-registered Hemangeol target in 2014 and the global standard.[6][7]

How IH patients come to harm — the preventable list

• Missing subglottic IH that presents as biphasic stridor at 4–8 weeks — an airway emergency needing urgent propranolol and ENT.[1]
• A periocular IH left without urgent ophthalmology and propranolol — permanent amblyopia and astigmatism from visual axis obstruction.[3]
• A large segmental facial IH without a PHACES screen — missed cerebrovascular anomalies, coarctation and stroke risk.[1][2]
• Five or more cutaneous IH without a liver ultrasound — missed hepatic haemangiomatosis and high-output cardiac failure.[3]
• Attributing Kasabach-Merritt to IH and treating with propranolol when the lesion is a KHE needing sirolimus.[4]
• Propranolol given without feeding guidance — masked hypoglycaemia during intercurrent illness.[9]
• Five or more cutaneous IH without a thyroid check — consumptive hypothyroidism from hepatic IH missed.[3]

Prognosis

Excellent for most IH — near-zero mortality in uncomplicated cases, with spontaneous involution the rule. About 20–40 percent retain residual telangiectasia, atrophy or fibrofatty tissue after complete involution, addressable with pulsed dye laser or surgery in later childhood. The psychosocial burden of a facial IH is real and justifies early treatment of disfiguring lesions, because the cosmetic outcome is best when propranolol is started in the early proliferative phase.[1][2]

The mantra, and the memory device

[1] [11]

The mantra: it is the commonest tumour of infancy and it mostly goes away — GLUT1-positive, absent at birth, proliferates then involutes. Watch the low-risk lesion; propranolol the high-risk one early, and never forget the airway, the eye, the PHACES screen, the liver, and that Kasabach-Merritt is KHE, not IH.[1][3]

Ward-round test — three stems, thirty seconds each

Stem 1 — the mushrooming cheek spot (answer)

A 6-week-old has a bright red, raised, lobulated plaque on the cheek that was barely visible at birth and has grown rapidly over three weeks. What is it, and what is the plan? Model: This is a superficial infantile haemangioma in its proliferative phase — absent or a precursor at birth, then rapid growth, is the signature; GLUT1-positive. It is not a vascular malformation (those are present and stable at birth). The plan is active non-intervention with photographic monitoring every 1–3 months through proliferation, and reassurance that about 50 percent involute by age 5 and 90 percent by 9. Treat with propranolol only if it develops a high-risk feature — ulceration, periocular position, airway, or significant disfigurement.[1][6]

Stem 2 — stridor at 6 weeks with a beard-distribution IH (answer)

A 6-week-old with a segmental haemangioma over the beard area develops biphasic stridor. What is happening, and what is the move? Model: This is a subglottic haemangioma — the beard (segmental) distribution IH is the surface clue, and biphasic stridor at 4–8 weeks is the airway emergency. The move is urgent propranolol escalated to 3 mg/kg/day plus ENT and anaesthesia review and bronchoscopy. Do not wait — subglottic IH can obstruct the airway rapidly. Initiate propranolol in a monitored setting.[1][3]

Stem 3 — the large segmental facial IH and the screen (answer)

A 2-month-old has a large segmental haemangioma covering over 5 cm of the forehead and scalp. What must you do beyond the skin? Model: Screen for PHACES syndrome — a segmental facial IH over 5 cm is the index lesion. Order MRI brain plus MRA, an echocardiogram, and an ophthalmology assessment to look for posterior fossa malformations, cerebrovascular and aortic arch anomalies, cardiac defects and eye anomalies. Before propranolol, complete the imaging and titrate the dose slowly because of the stroke risk from hypotension. A lumbosacral IH would instead trigger a spinal MRI for LUMBAR syndrome and a tethered cord.[1][2]

References

1. [1]Rodríguez Bandera AI, Sebaratnam DF, Wargon O, et al. Infantile hemangioma. Part 1: Epidemiology, pathogenesis, clinical presentation and assessment J Am Acad Dermatol, 2021.PMID 34419524
2. [2]Leung AKC, Lam JM, Leong KF, et al. Infantile Hemangioma: An Updated Review Curr Pediatr Rev, 2021.PMID 32384034
3. [3]Sebaratnam DF, Rodríguez Bandera AL, Wong LF, et al. Infantile hemangioma. Part 2: Management J Am Acad Dermatol, 2021.PMID 34419523
4. [4]Holm A, Mulliken JB, Bischoff J. Infantile hemangioma: the common and enigmatic vascular tumor J Clin Invest, 2024.PMID 38618963
5. [5]Kunimoto K, Yamamoto Y, Jinnin M. ISSVA Classification of Vascular Anomalies and Molecular Biology Int J Mol Sci, 2022.PMID 35216474
6. [6]Léauté-Labrèze C, Harper JI, Hoeger PH. Infantile haemangioma Lancet, 2017.PMID 28089471
7. [7]Léauté-Labrèze C, Voisard JJ, Moore N Oral Propranolol for Infantile Hemangioma N Engl J Med, 2015.PMID 26176392
8. [8]Léauté-Labrèze C, Dumas de la Roque E, Hubiche T, et al. Propranolol for severe hemangiomas of infancy N Engl J Med, 2008.PMID 18550886
9. [9]Biesbroeck L, Brandling-Bennett HA Propranolol for infantile haemangiomas: review of report of a consensus conference Arch Dis Child Educ Pract Ed, 2014.PMID 24242336
10. [10]Painter SL, Hildebrand GD. Review of topical beta blockers as treatment for infantile hemangiomas Surv Ophthalmol, 2016.PMID 26408055
11. [11]Cheng J, et al. Clinical characteristics, treatment outcomes and prognostic factors of ulcerated infantile hemangioma: 15 years of experience from a pediatric dermatology center in Hong Kong J Dermatolog Treat, 2025.PMID 41221590
12. [12]Maguiness SM, Hoffman WY, McCalmont TH, Frieden IJ. Early white discoloration of infantile hemangioma: a sign of impending ulceration Arch Dermatol, 2010.PMID 21079059
13. [13]Bardo DME, Gill AE, Iyer RS, et al. ACR Appropriateness Criteria® Soft Tissue Vascular Anomalies: Vascular Malformations and Infantile Vascular Tumors (Non-CNS)-Child J Am Coll Radiol, 2024.PMID 38823953
14. [14]Kumar MG, Coughlin C, Bayliss SJ Outpatient use of oral propranolol and topical timolol for infantile hemangiomas: survey results and comparison with propranolol consensus statement guidelines Pediatr Dermatol, 2015.PMID 25556828
15. [15]Krowchuk DP, Frieden IJ, Mancini AJ, et al. Clinical Practice Guideline for the Management of Infantile Hemangiomas Pediatrics, 2019.PMID 30584062
16. [16]Ábarzúa-Araya A, Navarrete-Dechent CP, Heusser F, Retamal J, Zegpi-Trueba MS. Atenolol versus propranolol for the treatment of infantile hemangiomas: a randomized controlled study J Am Acad Dermatol, 2014.PMID 24656727
17. [17]Gupta A, Kureel SN, Pandey A, et al. Angiotensin-converting Enzyme Inhibitors: Can it be a Potential Treatment of Infantile Hemangioma J Indian Assoc Pediatr Surg, 2021.PMID 34385766