Bruising and petechiae: the bleeding signs of thrombocytopenia at presentation
also called petechiae · bruising · easy bruising · purpura · thrombocytopenia · bleeding tendency
When leukaemia crowds out the marrow the body stops making platelets, so the child bruises in odd places and develops pinpoint red spots that do not fade when pressed. Bruising is seen in about 52% of children with leukaemia and a bleeding tendency in about 38%, and in a baby who is not yet walking, unexplained bruises should always be investigated.
The same thing, for a scientist
Thrombocytopenia at presentation is a consequence of megakaryocytic displacement and is the sign with the most immediate mortality relevance, because intracranial haemorrhage is the leading cause of early death in paediatric hyperleukocytosis. In the childhood leukaemia presenting-features meta-analysis (33 cohorts, n = 3084), bruising was present in 52%, bruising/petechiae in 42% and bleeding tendency in 38%. Coagulopathy frequently coexists: disseminated intravascular coagulation is driven by blast lysis and by tissue-factor exposure, and in hyperleukocytosis cohorts cerebral leukostasis and coagulopathy together accounted for early death in 4.2% of ALL and 7.7% of AML patients, all with WBC >= 400 x 10^9/L. In an infant, the differential for unexplained bruising includes non-accidental injury and inherited bleeding disorders, so the full blood count is both the fastest way to the diagnosis and the fastest way to exclude the wrong one.
What the signs are, and how to tell them apart1 deeper
- Petechiae: pinpoint, non-blanching red-purple macules, typically on the trunk, face and, in an infant, over the head and behind the ears where pressure is applied during handling. They do not fade under a glass slide. They are capillary bleeds, not infiltrates - which is what distinguishes them from leukaemia cutis (
concept-leukemia-cutis-infant), whose lesions are palpable nodules. - Bruising / purpura: larger ecchymoses in sites that ordinary handling would not explain. In a pre-mobile infant this is always abnormal.
- Mucosal bleeding: gum ooze, epistaxis, prolonged bleeding after a heel prick or immunisation.
- The dangerous ones: intracranial haemorrhage (irritability, seizure, bulging fontanelle, altered consciousness) and gastrointestinal bleeding.
| Sign | Frequency | Population |
|---|---|---|
| Bruising | 52% | childhood leukaemia, 33 cohorts, n = 3084 1 |
| Bruising/petechiae (combined) | 42% | same 1 |
| Bleeding tendency | 38% | same 1 |
| Rash (non-specific, heterogeneous) | 35% | same 1 |
The differential that has to be got right
Unexplained bruising in a baby generates two urgent and very different hypotheses: haematological disease, and non-accidental injury. Both are real, neither excludes the other, and getting the order wrong has consequences for the family that go well beyond medicine. The resolution is fast and cheap: a full blood count and film (concept-blood-film-and-blood-count-leukemia) plus a coagulation screen, available in under an hour, will separate thrombocytopenia from an inherited bleeding disorder from normal haemostasis. This is one of the strongest practical arguments for a low threshold to check a blood count in an unwell infant.
Why this is the sign that can kill in the first 48 hours
Thrombocytopenia plus hyperleukocytosis plus coagulopathy is the combination that produces early death. In a paediatric hyperleukocytosis series, intracranial haemorrhage rather than renal failure was the main cause of early death, cerebral leukostasis or coagulopathy-related death within 14 days occurred in 4.2% of ALL and 7.7% of AML patients, and every one of those patients had a white cell count of at least 400 x 10^9/L 2. The clinical corollary is asymmetric transfusion policy: platelets are given liberally and to a high threshold, while red cells are given cautiously because raising haematocrit raises viscosity (see concept-hyperleukocytosis-emergency). Fibrinogen and coagulation factors are replaced actively if disseminated intravascular coagulation is present.
Platelet recovery as the marker of a real remission
Platelet count is also the most robust simple marker of marrow recovery, which is why it appears in the definition of complete remission and in the criteria that permit the next chemotherapy block to start. In infant ALL that timing matters a great deal, because the induction that kills these babies is an infective complication during prolonged cytopenia: COG AALL0631 reduced induction mortality from 4/26 (15.4%) to 2/123 (1.6%) partly by keeping infants hospitalised through induction until there was evidence of marrow recovery 3. Counting platelets is therefore not bookkeeping; it is one of the decisions that determined whether these children lived.
| sign | bruising and petechiae secondary to thrombocytopenia |
|---|---|
| mechanism | megakaryocytic displacement by leukaemic marrow infiltration, often with superimposed disseminated intravascular coagulation from blast lysis |
frequencies — the numbers behind the chart
[
{
"finding": "bruising",
"pct": 52,
"population": "childhood leukaemia 0-18 y, n=3084"
},
{
"finding": "bruising/petechiae",
"pct": 42,
"population": "childhood leukaemia 0-18 y, n=3084"
},
{
"finding": "bleeding tendency",
"pct": 38,
"population": "childhood leukaemia 0-18 y, n=3084"
}
]dangerous manifestations
[ "intracranial haemorrhage", "gastrointestinal haemorrhage", "pulmonary haemorrhage" ]
early death from haemorrhage or leukostasis PCT
{
"ALL": 4.2,
"AML": 7.7,
"condition": "hyperleukocytosis, all with WBC >= 400 x 10^9/L"
}management
[ "liberal platelet transfusion to a high threshold", "cautious red-cell transfusion", "fibrinogen/factor replacement if DIC", "avoid intramuscular injections and arterial punctures" ]
differential in infancy
[ "non-accidental injury", "inherited platelet disorder", "von Willebrand disease", "immune thrombocytopenia", "sepsis with DIC", "vitamin K deficiency bleeding" ]
- Any agent that lyses a large blast burden quickly can precipitate or worsen disseminated intravascular coagulation at the moment the platelet count is at its lowest, so fast response and haemorrhagic death are coupled risks.
- Thrombocytopenia constrains the diagnostic procedures a targeted-therapy trial depends on: repeated marrow aspirates and lumbar punctures in a bleeding infant carry real risk, which limits how densely MRD and pharmacodynamics can be sampled.
- Because bruising is also the cardinal sign of non-accidental injury, delays and diagnostic detours at presentation are not purely medical, and any programme that depends on early diagnosis has to account for them.
- What proportion of KMT2A-rearranged infants have overt coagulopathy at presentation, and does it differ by fusion partner or by lineage (ALL versus AML, where DIC is classically more prominent)?
- Would earlier, lower-intensity cytoreduction with a targeted agent reduce the incidence of intracranial haemorrhage in infants presenting with a white cell count above 400 x 10^9/L, where haemorrhage is the dominant cause of early death?
- Does the pre-diagnostic interval in infant leukaemia differ between families whose child presents with visible bruising and those presenting with feeding problems alone, and does that translate into a difference in presenting white-cell count and risk group?
- 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 childhood leukaemia cohort studies (n = 3084, ages 0-18), bruising was present in 52% of children, bruising/petechiae in 42%, bleeding tendency in 38% and rash in 35%; hepatomegaly (64%), splenomegaly (61%), pallor (54%) and fever (53%) were the other features present in more than half or near half of children, and 6% 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
Cohort studies without control groups, so these are frequencies among diagnosed children rather than predictive values. 'Bruising', 'bruising/petechiae' and 'bleeding tendency' are overlapping categories recorded differently across the included studies, so they cannot simply be added. Ages 0-18 pooled without an infant stratum; bruising is likely to be noticed differently in a pre-mobile infant than in a toddler, in both directions.
Evidence that argues against this (1)
- These frequencies are a poor basis for suspecting leukaemia in an individual child, and quoting them without the base rate inverts their meaning. They are prevalences measured in children who ALREADY had leukaemia - 52% bruising is P(sign | disease), not P(disease | sign). Bruising is close to universal in healthy active children, and childhood leukaemia is rare, so the positive predictive value of bruising alone is very low. The meta-analysis also pools 33 cohorts across ages 0-18 and both lineages, and infant KMT2A-rearranged presentation - hyperleukocytosis, organomegaly, CNS involvement - is not the modal presentation these numbers describe. Archives of Disease in Childhood 2016
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In a paediatric acute leukaemia cohort with hyperleukocytosis (48 ALL, 13 AML) intracranial haemorrhage rather than renal failure was the main cause of early death; cerebral leukostasis or coagulopathy-related death within the first 14 days occurred in 4.2% of ALL and 7.7% of AML patients, and all of these patients had a white cell count of at least 400 x 10^9/L; laboratory tumour lysis developed in 54.16% of ALL and 76.92% of AML patients with no patient requiring dialysis.
Management of Hyperleukocytosis in Childhood Acute Leukemia Without Leukapheresis and Rasburicase Prophylaxis. · Aylan Gelen S, Sarper N, Zengin E, et al. · Journal of Pediatric Hematology/Oncology 2022What this evidence cannot show
Single-centre retrospective series of 61 children, not infants specifically; leukapheresis was unavailable rather than withheld by design. Early-death percentages rest on two or three events each, so the difference between the ALL and AML figures is not meaningful. The study does not report platelet counts or transfusion thresholds, so it supports the claim about cause of death but not a specific transfusion policy.
Evidence that argues against this (1)
- This cuts against reading the entry as being about platelets. In hyperleukocytosis the lethal event was intracranial haemorrhage driven by cerebral leukostasis or coagulopathy, which is a white-cell and clotting problem rather than a simple consequence of marrow crowding. Correcting the platelet count does not address it, and a framing built on 'marrow crowded out, so platelets fall' will mis-triage exactly the children at highest early risk. Journal of Pediatric Hematology/Oncology 2022
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In COG AALL0631, de-intensifying induction and adding enhanced supportive care including hospitalisation through induction until evidence of marrow recovery reduced induction mortality from 4 of 26 (15.4%) to 2 of 123 (1.6%) (p = 0.009) and sterile-site infections from 15 of 26 (57.7%) to 24 of 123 (19.5%) (p = 0.0002), while complete response at end of induction intensification rose from 17 of 25 (68%) to 94 of 100 (94%) (p = 0.0012).
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
Sequential cohorts rather than a randomised comparison, and the first cohort was small (n = 26), so 15.4% rests on 4 deaths; the change bundled a less intensive regimen with a supportive-care package, so the individual contributions cannot be separated. The improvement in complete response rate is also confounded by a different assessment timepoint (week 9 versus week 7).
Evidence that argues against this (1)
- The AALL0631 improvement cannot be attributed to bleeding or platelet management. The modification bundled de-intensified induction WITH enhanced supportive care and hospitalisation through marrow recovery, so the mortality reduction is a package effect. Infection during a de-intensified induction is at least as plausible a mediator as haemorrhage, and the design does not separate them. The denominators are also very small. Pediatric Blood & Cancer 2015
displacement of megakaryocytes by blasts produces the presenting thrombocytopenia
presenting sign: thrombocytopenic bleeding from marrow replacement
- co-occurs with →Hepatosplenomegaly: the swollen abdomen that is often the first thing a parent notices
both follow from the same process: blast expansion displacing normal haematopoiesis and seeding organs
multiple petechiae with INR 7.7 and a large intracranial haemorrhage
cutaneous haemorrhages with grade 3 disseminated intravascular coagulation at presentation
spontaneous gum bleeding with platelets 117 x 10^9/L at presentation