Rash: the one presenting feature that gets commoner the younger the baby is
also called skin infiltration infant leukaemia · rash at presentation infant · blueberry muffin leukaemia · leukaemia cutis frequency by age
Rash is listed at 35% in the pooled childhood figures, but in leukaemia the skin lesions that matter are leukaemia cells growing in the skin, and those cluster in the very youngest babies. In a 144-infant series every recorded case of skin infiltration was in a baby under six months old.
The same thing, for a scientist
The pooled 0-18 figure lists rash at 35% [E1], a category that mixes petechial rash from thrombocytopenia with true leukaemic skin infiltration. The infant-specific signal is about the second. In a 144-case infant leukaemia series, skin infiltration was recorded in six patients and **all six were under 6 months of age** [E2]. Neonatal leukaemia is characterised by leukaemia cutis as one of a triad with hepatosplenomegaly and hyperleucocytosis [E3]. A 2026 neonatal case presented with generalised purpuric 'blueberry muffin' lesions alongside a leucocyte count of 615,400/microlitre [E4]. The gradient runs the opposite way to most presenting features: rather than the infant figure being an extrapolation from older children, this is a sign whose frequency rises steeply as age falls, and whose peak is in the neonatal window where the pooled figure has essentially no data at all. The pooled category is also the wrong category, because 'rash' at 35% and 'bruising/petechiae' at 42% are both in the same list [E1] and a petechial rash is a platelet sign, not a skin-infiltration sign.
Two different things called rash1 deeper
| Finding | Mechanism | What it implies |
|---|---|---|
| Petechiae and purpura | thrombocytopenia from marrow replacement | a platelet count, urgently |
| Leukaemia cutis: firm violaceous or blue-grey nodules | leukaemic cells infiltrating dermis | extramedullary disease; strongly associated with neonatal and infant presentation and with monocytic/myeloid phenotype |
| Blueberry-muffin lesions | dermal haematopoiesis or infiltration | classically congenital infection first, then congenital leukaemia and neuroblastoma |
The pooled meta-analysis reports 'rash 35%' and 'bruising/petechiae 42%' as separate ranked entries 1 without saying whether infiltrative lesions were distinguished from haemorrhagic ones in the source cohorts. For an infant, they are different emergencies.
The age gradient inside the first year
Six of 144 infants (4.2%) had skin infiltration, and the paper states all six were under 6 months 2. Treating the under-six-month subgroup as the denominator is not possible from the abstract, because the paper reports that 38.2% of the cohort were aged 9-12 months but does not give the size of the under-six-month group. If the under-six-month group were, say, 25% of 144 (36 infants), 6/36 would be 17% - roughly four times the whole-cohort rate. That arithmetic is illustrative, not a result, and it is written out here so that nobody mistakes it for one.
Why this matters for a diagnostic pathway
A neonate with skin lesions is triaged to congenital infection - TORCH serology, and correctly so, because congenital infection is far commoner. Neonatal leukaemia sits behind that screen 3, and the thing that brings it forward is the full blood count, not the skin. The clinical rule that follows is narrow and checkable: a neonate with blueberry-muffin or nodular skin lesions needs a full blood count and film on the same visit as the TORCH screen, because the alternative diagnosis is one where hours matter.
| metric | presenting sign frequency by age stratum |
|---|---|
| population | infants <1 year with acute leukaemia |
pooled childhood
{
"rash_pct": 35,
"bruising_petechiae_pct": 42,
"n": 3084,
"pmid": "27647842",
"title": "Clinical presentation of childhood leukaemia: a systematic review and meta-analysis.",
"authors": "Clarke RT, Van den Bruel A, Bankhead C, et al.",
"venue": "Arch Dis Child",
"year": 2016,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
}infant observation
{
"cohort": "single centre China",
"n": 144,
"skin_infiltration_n": 6,
"all_under_6_months": true,
"pmid": "26714909",
"title": "[Clinical analysis of 144 cases of infant leukaemia].",
"authors": "Wen X, Guan X, Xian Y, et al.",
"venue": "Nan Fang Yi Ke Da Xue Xue Bao",
"year": 2015,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
}- What is the frequency of leukaemia cutis within the under-three-month stratum of a modern infant ALL trial, separated from AML? No trial reports it.
- Is the age gradient in skin infiltration a lineage gradient in disguise, given that younger infants more often have myeloid or mixed-phenotype disease?
- How long is the interval between a neonate's skin lesions being noticed and a full blood count being taken, and does shortening it change outcome?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review
In the pooled 0-18 meta-analysis rash was reported in 35% and bruising/petechiae in 42% of children with leukaemia.
Clinical presentation of childhood leukaemia: a systematic review and meta-analysis. · Clarke RT, Van den Bruel A, Bankhead C, Mitchell CD, Phillips B, Thompson MJ · Arch Dis Child 2016What this evidence cannot show
The two categories are ranked separately but the review does not state whether the source cohorts distinguished infiltrative from haemorrhagic lesions, so the 35% is of uncertain composition. Ages 0-18 pooled.
Evidence that argues against this (1)
- Within infancy the skin sign is concentrated in the under-six-month group, so a single pooled percentage across 0-18 both under-states the neonatal rate and over-states the rate in an 11-month-old. Wen X et al. 2015
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In a 144-case single-centre infant leukaemia series, six patients had skin infiltration and all six were under 6 months of age.
[Clinical analysis of 144 cases of infant leukaemia]. · Wen X, Guan X, Xian Y, Dou Y, Guo Y, Xiao J, et al. · Nan Fang Yi Ke Da Xue Xue Bao 2015What this evidence cannot show
Six events. The size of the under-six-month subgroup is not given in the abstract so no rate within that subgroup can be computed. Mixed lineage series (83 ALL, 55 AML, 1 hybrid, 5 unclassifiable) and leukaemia cutis is commoner in myeloid disease, so lineage is confounded with age here. Chinese-language paper; English abstract read, full text not retrieved.
Evidence that argues against this (1)
- Because the series contains 55 AML cases and leukaemia cutis is a predominantly monocytic/myeloid phenomenon, the age gradient could be a lineage gradient: younger infants are more often myeloid or mixed-phenotype. van der Linden MH et al. 2012
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.review
Neonatal leukaemia typically presents with hepatosplenomegaly, leukaemia cutis and/or hyperleucocytosis, and congenital infection must be ruled out before the diagnosis is made.
Diagnosis and management of neonatal leukaemia. · van der Linden MH, Creemers S, Pieters R · Semin Fetal Neonatal Med 2012What this evidence cannot show
Narrative review; 'typically' is not a measured frequency and no denominator is given. Neonatal AML is commoner than neonatal ALL, so the triad is weighted towards myeloid disease and may not describe KMT2A-rearranged infant ALL specifically.
Evidence that argues against this (1)
- The Interfant-99 congenital ALL analysis, the largest uniformly treated congenital series, characterises congenital disease by higher white cell count and a trend to more MLL rearrangement and CD10-negative B-lineage disease, and does not lead with skin involvement. van der Linden MH et al. 2009
- E4weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
A neonate with mixed phenotype acute leukaemia presented at 7 days with generalised purpuric 'blueberry muffin' skin lesions together with a peak leucocyte count of 615,400/microlitre, severe anaemia and thrombocytopenia.
Extreme hyperleukocytosis and blueberry muffin skin lesions as the initial presentation of neonatal leukemia: a case report. · Bustamante-Ordonez MA, Macavilca-Toribio MA, Arenas-Vasquez JR, Roman-Lazarte V, Zavaleta-Corvera C, Samanez-Obeso A · BMC Pediatr 2026What this evidence cannot show
Single case report; establishes co-occurrence, not frequency. KMT2A testing was negative, so this is not the biology this graph centres on. Cited for the clinical sign only; no narrative about the patient is reproduced here.
Evidence that argues against this (1)
- Blueberry-muffin lesions are far more often congenital infection or neuroblastoma than leukaemia, so their positive predictive value for leukaemia in an unselected neonate is low and a case series cannot estimate it. van der Linden MH et al. 2012
age gradient within infancy for skin infiltration
presenting skin signs in the first year of life
- ← related toLeukaemia diagnosed in the first month of life: a different presentation and a different prognosis
the skin sign peaks in the neonatal window