Hepatosplenomegaly: the swollen abdomen that is often the first thing a parent notices
also called hepatomegaly · splenomegaly · organomegaly · big liver and spleen · distended abdomen
In leukaemia the liver and spleen fill up with leukaemia cells and grow large enough to feel through the abdominal wall, which is why a baby's tummy can look swollen and feel firm. Across childhood leukaemia a palpable liver is found in about 64% of children and a palpable spleen in about 61%, and in infants the organs tend to be bigger still.
The same thing, for a scientist
Hepatosplenomegaly in acute leukaemia reflects leukaemic infiltration superimposed on the fetal and neonatal liver's residual capacity for extramedullary haematopoiesis. A meta-analysis of 33 cohort studies (n = 3084 children, 0-18 years) ranks hepatomegaly (64%) and splenomegaly (61%) as the two most frequent presenting signs of childhood leukaemia, with hepatosplenomegaly recorded as a combined finding in 42%. In infant and congenital disease the frequency and degree are greater: hepatosplenomegaly was reported in 47.5% of published congenital leukaemia cases, which is a lower bound given the acuity of that population. Clinically, massive organomegaly in an infant compromises diaphragmatic excursion and feeding, raises the risk of splenic rupture, and is a marker of total tumour burden that travels with hyperleukocytosis and tumour-lysis risk.
Why the liver and spleen, and why especially in a baby1 deeper
Two mechanisms stack.
The first is simple infiltration: the spleen filters blood and the liver receives the portal circulation, so both are exposed to any circulating cell population and both have sinusoidal architecture that traps large, stiff blasts. When the count is in the hundreds of thousands (see concept-hyperleukocytosis-emergency), organ infiltration is essentially guaranteed.
The second is developmental and specific to infancy. The fetal liver is a haematopoietic organ, and KMT2A-rearranged infant leukaemia is thought to originate in a fetal-liver-stage progenitor (cell-type-fetal-liver-hsc). A leukaemia of fetal origin is therefore expanding in a tissue that recently supported normal blood formation and has not fully lost the niche signals for it. That is a mechanistic prediction, not an established cause of the organomegaly, but it is testable and it explains why the neonatal and infant presentation is so organomegaly-heavy.
What the numbers actually are
| Finding | Frequency | Population |
|---|---|---|
| Hepatomegaly | 64% | childhood leukaemia, 33 cohorts, n = 3084 1 |
| Splenomegaly | 61% | same 1 |
| Hepatosplenomegaly (recorded as one finding) | 42% | same 1 |
| Abdominal distension | reported as a common presenting feature | same 1 |
| Hepatosplenomegaly | 47.5% | published congenital leukaemia cases 2 |
The honest caveat is that no meta-analysis of presenting signs restricted to infants under 1 year exists, so the 64%/61% figures are for children 0-18 years and the infant-specific figure comes from a case-report literature with strong publication bias. The direction of the bias is at least predictable: case reports over-select dramatic presentations, while a mixed-age meta-analysis dilutes infants with adolescents whose organomegaly is less prominent.
What it means for the baby in front of you1 deeper
Massive organomegaly in a 5 kg infant is not a cosmetic finding. It splints the diaphragm and adds to the respiratory compromise of pulmonary leukostasis and anaemia; it reduces gastric capacity so the baby feeds poorly and is often described by parents as 'off feeds' or 'not finishing bottles' before anyone notices the abdomen; it puts the spleen at risk of traumatic rupture, which is why handling and car-seat positioning are discussed with parents; and it is a proxy for total tumour burden, so it predicts the tumour-lysis load once treatment starts.
It is also the reason the pre-diagnosis story is so often a gastrointestinal one. Because 6% of children with leukaemia are asymptomatic at diagnosis and the rest present with non-specific features such as pallor, fever and abdominal symptoms 1, the commonest route to diagnosis is repeated primary-care visits for feeding problems, irritability or a 'viral illness' until someone palpates the abdomen or checks a blood count.
The differential a clinician must exclude
In an infant, hepatosplenomegaly is not specific to leukaemia. The list includes storage disorders, congenital infection (CMV, toxoplasmosis, syphilis), haemolytic anaemia, neuroblastoma with hepatic metastases, juvenile myelomonocytic leukaemia, haemophagocytic lymphohistiocytosis and transient abnormal myelopoiesis in a child with trisomy 21. This is why the blood film and count (concept-blood-film-and-blood-count-leukemia) rather than the abdominal examination is the step that actually makes the diagnosis.
| sign | hepatosplenomegaly |
|---|---|
| mechanism | leukaemic infiltration of sinusoidal organs plus residual fetal-liver haematopoietic niche |
frequencies — the numbers behind the chart
[
{
"finding": "hepatomegaly",
"pct": 64,
"population": "childhood leukaemia 0-18 y, n=3084",
"source": "Clarke 2016 meta-analysis"
},
{
"finding": "splenomegaly",
"pct": 61,
"population": "childhood leukaemia 0-18 y, n=3084",
"source": "Clarke 2016 meta-analysis"
},
{
"finding": "hepatosplenomegaly (combined)",
"pct": 42,
"population": "childhood leukaemia 0-18 y, n=3084",
"source": "Clarke 2016 meta-analysis"
},
{
"finding": "hepatosplenomegaly",
"pct": 47.5,
"population": "published congenital leukaemia cases",
"source": "StatPearls"
}
]consequences
[ "respiratory compromise from diaphragmatic splinting", "poor feeding", "splenic rupture risk", "marker of total tumour burden and tumour-lysis load" ]
differential in infancy
[ "storage disorders", "congenital CMV/toxoplasmosis/syphilis", "haemolytic anaemia", "neuroblastoma with hepatic metastases", "juvenile myelomonocytic leukaemia", "HLH", "transient abnormal myelopoiesis" ]
- Organomegaly is a reservoir of disease in a compartment where drug exposure and immune-effector access differ from the marrow, so a response measured in blood or marrow may overstate systemic clearance.
- Rapid lysis of a heavily infiltrated liver and spleen releases a large metabolic load and can produce hepatic dysfunction that then impairs clearance of the very drug causing it - relevant for orally dosed, hepatically metabolised menin inhibitors.
- Because organomegaly is non-specific in infancy, a targeted therapy programme that relies on rapid diagnosis will still be limited by the pre-diagnostic interval spent investigating feeding problems and viral illness.
- No meta-analysis of presenting signs restricted to infants under 1 year exists; the Interfant and COG consortia hold that data in their trial databases and a simple descriptive analysis would close this gap.
- Does the degree of hepatomegaly at diagnosis correlate with a fetal-liver transcriptional signature in the blasts, as the cell-of-origin model would predict, or only with total tumour burden?
- Does organomegaly resolve faster under menin inhibition than under conventional induction, and could the rate of resolution serve as an early pharmacodynamic readout in a baby too small for frequent marrow sampling?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In a systematic review and meta-analysis of 33 cohort studies covering 3084 children aged 0-18 with leukaemia, five features were present in more than 50% of children - hepatomegaly 64%, splenomegaly 61%, pallor 54%, fever 53% and bruising 52% - and a further eight in a third to a half, including recurrent infections 49%, fatigue 46%, limb pain 43%, hepatosplenomegaly 42%, bruising/petechiae 42%, lymphadenopathy 41%, bleeding tendency 38% and rash 35%; 6% of children were asymptomatic at diagnosis.
Clinical presentation of childhood leukaemia: a systematic review and meta-analysis · Clarke RT, Van den Bruel A, Bankhead C, et al. · Archives of Disease in Childhood 2016What this evidence cannot show
All 33 included studies were cohort studies without control groups, so these are frequencies among children already diagnosed, not positive predictive values for a child presenting to primary care. The age range is 0-18 years and infants are not analysed separately, so the figures cannot be attributed to infant ALL specifically. Ascertainment of 'hepatomegaly' depends on clinical palpation, which in a small infant is both easier (thin abdominal wall) and less standardised than imaging.
Evidence that argues against this (1)
- This meta-analysis covers ages 0 to 18, and infants are a small minority of its 3,084 children. Infant KMT2A-rearranged leukaemia presents differently from the disease that dominates the pooled figure — higher white cell counts, more CNS and extramedullary involvement, a pro-B rather than common-ALL phenotype — so a frequency computed across the whole paediatric range is not the frequency in the patients this graph is about. The signs are also chart-recorded retrospectively, which systematically under-records soft findings and over-records the ones that prompted referral. Organomegaly is additionally examiner-dependent: a palpable liver edge is normal in infants, so the threshold for recording hepatomegaly differs between the infant and the older-child part of the pooled population in the direction that inflates it in infants. Statistics in Medicine 2006
- E2weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.book
In reported congenital leukaemia, hepatosplenomegaly occurred in 47.5% of cases alongside leukaemia cutis in 67.8%, CSF blasts in 25.4% and hyperleukocytosis in 62.7% with a mean WBC of 68.5 x 10^9/L; skin involvement was reported in 14.5% of children under two years versus 2.6% of older children.
Infant Leukemia (StatPearls) · StatPearls Publishing · NCBI Bookshelf 2024What this evidence cannot show
Tertiary source summarising pooled published congenital leukaemia case reports. Case-report literature massively over-represents visually striking presentations, so leukaemia cutis at 67.8% is almost certainly inflated; the hepatosplenomegaly figure of 47.5% is probably closer to a floor because organomegaly is less likely to be the reason a case gets written up. Treat as an order of magnitude only.
Evidence that argues against this (1)
- The mixed-age meta-analysis records hepatomegaly in 64% and splenomegaly in 61%, i.e. higher than the 47.5% congenital figure, which is the opposite of what a purely biological argument would predict and shows how much these numbers depend on how the finding was recorded. source 2016
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Infants with ALL are more likely than older children to have hyperleukocytosis, central nervous system disease and 11q23/KMT2A rearrangement, and intensification of therapy in this group reduced relapse but increased early deaths without improving survival.
Decreased induction morbidity and mortality following modification to induction therapy in infants with acute lymphoblastic leukemia enrolled on AALL0631: a report from the Children's Oncology Group · Salzer WL, Jones TL, Devidas M, et al. · Pediatric Blood & Cancer 2015What this evidence cannot show
This paper's subject is induction toxicity, not presenting signs; the statement about hyperleukocytosis and CNS disease is a framing claim in its introduction and it gives no prevalence percentages for its own cohort. Used here only to establish the qualitative contrast between infants and older children.
Evidence that argues against this (1)
- The entry reports the intensification result and does not draw its conclusion: therapy intensification reduced relapse but increased early deaths WITHOUT improving survival. That is the central clinical fact of this disease and it argues that presenting features which mark tumour burden cannot be acted on by giving more of the same treatment. Leukemia 2021
- ← refutesBig liver, big spleen: the organomegaly numbers measured in infants, not pooled from children
the pooled 64%/61% figures are estimated outside the infant population they are applied to
presenting sign: organ infiltration, often the finding that prompts the blood count
both follow from the same process: blast expansion displacing normal haematopoiesis and seeding organs
both follow from the same process: blast expansion displacing normal haematopoiesis and seeding organs
- ← co-occurs withFever and infection: the reason most of these babies see a doctor, and the reason some of them die
both follow from the same process: blast expansion displacing normal haematopoiesis and seeding organs
liver 4 cm, spleen 5 cm below costal margin
marked hepatosplenomegaly at birth; post-mortem showed diffuse sinusoidal and red-pulp infiltration
hepatosplenomegaly at presentation in a 4-month-old