Thrombocytopenia in pregnancy
White cells and platelets in pregnancy and the newborn: thrombocytopenia, ITP, TTP
Pregnancy and labour raise the white count and gently lower platelets, so many abnormal-looking results are normal. CPSP asks you to separate gestational thrombocytopenia from immune, pre-eclamptic, microangiopathic and alloimmune causes, and to explain each mechanism and its treatment.
One of the most tested topics in past papers
White cells in pregnancy and labour
- The white count rises through pregnancy (up to about 15 × 10^9/L) and in labour reaches 20–30 × 10^9/L, almost all neutrophils. Cortisol and adrenaline demarginate neutrophils from vessel walls; it settles within a week. Judge infection clinically, not on the count alone.
- Corticosteroids (antenatal betamethasone): neutrophils rise; eosinophils, lymphocytes and monocytes fall.
- Neutrophil alkaline phosphatase is high in pregnancy and reactive leucocytosis but low in CML.
- Infection: band forms, toxic granulation and Döhle bodies (left shift). In chorioamnionitis, neutrophil matrix metalloproteinases and elastase digest amniotic collagen, causing preterm rupture of membranes.
How neutrophils reach infection
- Sequence: margination, rolling, firm adhesion, transmigration, then chemotaxis and phagocytosis.
- Rolling on P-selectin (from Weibel–Palade bodies) and E-selectin, induced early by TNF and IL-1; adhesion by beta-2 integrins (CD18) to ICAM-1; transmigration (diapedesis) between endothelial cells by PECAM-1 (CD31); chemotaxis towards IL-8, C5a and LTB4.
- Killing uses the respiratory burst: NADPH oxidase makes superoxide, and myeloperoxidase makes hypochlorite. Its absence is chronic granulomatous disease.
- CD18 defect (leucocyte adhesion deficiency): delayed cord separation, infection without pus, neutrophilia.
White cells in the newborn
- Day 1 count is 10–26 × 10^9/L; lymphocytes dominate by the end of week 1.
- Neonatal sepsis: an immature to total neutrophil (I:T) ratio above 0.2 suggests infection. The marrow storage pool is small, so severe sepsis causes neutropenia, a poor sign.
- Babies of mothers with severe pre-eclampsia and growth restriction have mild neutropenia and thrombocytopenia from reduced production caused by chronic fetal hypoxia (marrow favours red cell production via erythropoietin). Cultures are negative and the mother's count is normal; recovery takes about 10 days, but the neutropenia still raises infection risk.
- Alloimmune neonatal neutropenia: maternal IgG against paternal neutrophil (HNA) antigens.
- Transient abnormal myelopoiesis in Down syndrome newborns (GATA1) resolves by 3 months, with later AML risk.
Platelets in pregnancy
- The count falls slightly in normal pregnancy (haemodilution and faster clearance). Von Willebrand factor and fibrinogen rise, so mild von Willebrand disease may be masked at term but levels fall after delivery, causing secondary PPH.
- Always check the film: EDTA-dependent clumping gives a falsely low count with no bleeding.
| Cause | Timing and features | Mechanism | Baby and other notes |
|---|---|---|---|
| Gestational thrombocytopenia (commonest, about three-quarters) | Late pregnancy; usually above 70 × 10^9/L; no history; normal BP and film | Haemodilution plus increased clearance | Normal count |
| Immune thrombocytopenia (ITP) | Any time, often first trimester, may be below 50; history of bruising; large platelets; marrow shows more megakaryocytes | IgG against GPIIb/IIIa; coated platelets removed by splenic macrophages via Fc receptors | IgG crosses via FcRn; neonatal nadir days 2–5 |
| Pre-eclampsia and HELLP | After 20 weeks with hypertension; RUQ pain, raised ALT and LDH | Platelet consumption on damaged endothelium; schistocytes | PT, APTT and fibrinogen stay normal until DIC sets in |
| DIC | Abruption, amniotic fluid embolism, sepsis, retained dead fetus | Decidual tissue factor consumes platelets and fibrinogen; raised D-dimer | PT and APTT prolonged |
| TTP | Fever, confusion, petechiae, schistocytes, renal signs, normal PT and APTT | ADAMTS13 below 10% (usually an autoantibody inhibitor) leaves ultra-large vWF multimers that string platelets into microthrombi | Plasma exchange; avoid platelet transfusion |
| HUS | Bloody diarrhoea (Shiga toxin *E. coli*) in children; atypical form postpartum with renal failure | Toxin or uncontrolled alternative complement damages glomerular endothelium | ADAMTS13 normal |
Treating immune and microangiopathic thrombocytopenia
- ITP in pregnancy: treat if platelets are below about 20–30 × 10^9/L, there is bleeding, or delivery approaches (aim above 50 for delivery, about 80 for neuraxial anaesthesia). First line is prednisolone or IVIG. If bleeding continues despite steroids, give intravenous immunoglobulin: it saturates Fc receptors on splenic macrophages and FcRn, so coated platelets escape destruction and the count rises in 1–3 days. Platelet transfusion is futile except for life-threatening bleeding or surgery, because the antibody destroys transfused platelets too.
- TTP: daily plasma exchange removes the anti-ADAMTS13 inhibitor and the large vWF multimers, and the donor plasma replaces ADAMTS13. Platelets feed the microthrombi and can worsen thrombosis.
Alloimmune and drug-related thrombocytopenia
- Neonatal alloimmune thrombocytopenia: a well baby with severe thrombocytopenia and a mother with a normal count. The mother lacks HPA-1a, carried on glycoprotein IIIa; the fetus inherits it from the father. It often affects the first pregnancy. Platelets carry HLA class I, not class II, antigens.
- Heparin-induced thrombocytopenia: day 5–10 (for example 10 days after cardiac surgery), IgG against heparin–platelet factor 4, causing thrombosis rather than bleeding.
- Low-dose aspirin irreversibly acetylates COX-1. Platelets have no nucleus and cannot make new cyclo-oxygenase, so thromboxane A2 stays low for their 7–10 day life, while nucleated endothelium makes new enzyme and recovers prostacyclin within hours. Pre-eclampsia shows high thromboxane and low prostacyclin, hence aspirin from 12 weeks.
- Reactive thrombocytosis: after haemorrhage, surgery, delivery or splenectomy and with iron deficiency.
Inherited and marrow disorders
| Disorder | Defect | Clue |
|---|---|---|
| Bernard–Soulier | GPIb (vWF receptor) | Giant platelets, low count, no ristocetin response; menorrhagia, gum bleeding |
| Glanzmann thrombasthenia | GPIIb/IIIa | Normal count, no aggregation except with ristocetin |
| Von Willebrand disease | vWF | Long bleeding time, normal platelet count, APTT may be long |
| Wiskott–Aldrich syndrome | WAS gene, X-linked | Eczema, small platelets, recurrent infections in a boy |
| Polycythaemia vera | JAK2 V617F | Itch after hot baths, plethora, splenomegaly, high Hb, white cells and platelets |
| Megaloblastic anaemia | Faulty DNA synthesis | All precursors affected: pancytopenia with hypersegmented neutrophils |
- Vitamin K deficiency (long antibiotics with parenteral nutrition) and warfarin prolong the PT with a normal platelet count; a brother who bled after circumcision suggests haemophilia (factor VIII or IX assay).
Remember Mild, late and well: gestational thrombocytopenia. Early, severe or with a history: think ITP.
Exam trap A normal maternal count with a severely thrombocytopenic newborn points to alloimmune thrombocytopenia, not maternal ITP.
Exam trap Normal clotting tests do not exclude HELLP or TTP; they only show that DIC has not set in.
How FCPS asks about this
- Fever, confusion, low platelets, schistocytes, normal PT and APTT: TTP; treat with plasma exchange.
- ITP at 32 weeks bleeding despite steroids: intravenous immunoglobulin, not platelets.
- Antigen in severe alloimmune thrombocytopenia: HPA-1a on GPIIIa.
- Why low-dose aspirin spares prostacyclin: platelets cannot make new COX.
- Sequence of neutrophil events in acute inflammation: margination, rolling, adhesion, transmigration.
- Young woman with menorrhagia and giant platelets: Bernard–Soulier syndrome.
- Eczema, small platelets, infections in a boy: Wiskott–Aldrich syndrome.
Test yourself on this topic
5 of the 82 questions in our bank on this topic. Choose an answer to see why every option is right or wrong.
Six hours after a normal vaginal delivery, a woman's white cell count is 24 × 10⁹/L, almost all neutrophils. Her temperature is 37.1 °C, the uterus is firm and non-tender and the lochia is not offensive. Which mechanism best explains this count?
Answer: Demargination by cortisol and adrenaline
Labour is intense physical stress. Cortisol and adrenaline detach neutrophils that are loosely stuck to vessel walls (the marginated pool) and release stored neutrophils from the marrow, so the count can reach 20–30 × 10⁹/L in labour and the first postpartum day. It settles within about a week. A raised count alone is therefore a poor sign of infection after delivery; fever, uterine tenderness and offensive lochia matter more. Haemoconcentration raises all cells only slightly.
Labour leucocytosis
- WBC up to 20–30 × 10⁹/L in labour
- Stress hormones demarginate neutrophils
- Settles within about a week
- Judge infection on clinical signs
Remember High WBC after delivery is not sepsis by itself.
Exam trap Antibiotics should not be started for a postpartum white count alone.
Why the other options are wrong
- Endotoxin release from the placental bed No sign of infection; endotoxin is not released in normal delivery.
- Fetal white cells entering her circulation Fetal leucocytes do not enter maternal blood in such numbers.
- Oxytocin stimulating granulocyte production Oxytocin has no effect on granulopoiesis.
- Haemoconcentration from dehydration Dehydration raises counts only slightly and proportionally.
A woman at 34 weeks with pyelonephritis has a white cell count of 32 × 10⁹/L with neutrophils, bands and some myelocytes. There is no splenomegaly and platelets are normal. Which result most strongly favours a reactive cause over chronic myeloid leukaemia?
Answer: High neutrophil alkaline phosphatase
In infection, pregnancy and other reactive states, neutrophils contain plenty of alkaline phosphatase, so the leucocyte (neutrophil) alkaline phosphatase score is high; pregnancy itself raises it. In CML the clonal neutrophils have a low score. CML also shows basophilia, eosinophilia, a peak of myelocytes over metamyelocytes and a high serum B12 from increased transcobalamin made by granulocytes. Today BCR-ABL1 testing confirms the diagnosis, but the NAP score still illustrates the difference.
Reactive vs CML
- Reactive: NAP high
- CML: NAP low
- CML: basophils, myelocyte peak
- Pregnancy raises NAP
Remember High NAP: reactive or pregnant, not CML.
Exam trap Pregnancy raises the NAP score, so a high score cannot be blamed on infection alone.
Why the other options are wrong
- Raised absolute basophil count Basophilia favours CML.
- More myelocytes than metamyelocytes A myelocyte peak favours CML.
- Very high serum vitamin B12 level High B12 from transcobalamin favours CML.
- Eosinophilia together with basophilia Eosinophilia with basophilia favours CML.
In pre-eclampsia, changes in the placenta and platelets favour vasoconstriction, platelet aggregation and placental thrombosis. Which imbalance of prostanoids is characteristic?
Answer: Thromboxane A2 high, prostacyclin low
In normal pregnancy prostacyclin production by endothelium and trophoblast rises, helping to keep vessels dilated and platelets quiet. In pre-eclampsia the damaged endothelium makes less prostacyclin, while activated platelets and the placenta make more thromboxane A2, which constricts vessels and promotes aggregation. This shift is the basis for using low-dose aspirin, which suppresses platelet thromboxane. Leukotrienes and lipoxins are not the central pathway.
Prostanoid balance
- Normal pregnancy: prostacyclin up
- Pre-eclampsia: thromboxane A2 up
- Prostacyclin down
- Basis of low-dose aspirin
Remember Pre-eclampsia tips the balance to thromboxane.
Exam trap Do not confuse the normal pregnancy rise in prostacyclin with the pre-eclampsia picture.
Why the other options are wrong
- Prostacyclin high, thromboxane A2 low That is the normal pregnancy pattern, opposite to pre-eclampsia.
- Prostaglandin E2 high, PGF2α low These are not the key pair in pre-eclampsia.
- Leukotriene B4 high, lipoxin A4 low Leukotrienes and lipoxins are not the central change.
- Prostacyclin high, PGE2 low Prostacyclin falls in pre-eclampsia.
Ten days after a postpartum haemorrhage of 2.5 litres treated with transfusion and repair of a cervical tear, a woman's platelet count is 640 × 10⁹/L. Hb is 9.5 g/dL and the white cell count is normal. What is the most likely cause?
Answer: Reactive response to bleeding
After haemorrhage, surgery and delivery itself, the platelet count commonly rises over 1–3 weeks. Inflammatory cytokines, especially interleukin-6, raise hepatic thrombopoietin production and drive megakaryocytes; iron deficiency from blood loss adds to the rise. Reactive thrombocytosis rarely causes thrombosis by itself and settles as the stimulus resolves. Essential thrombocythaemia and CML are clonal disorders that need a persistent count and other features to suspect.
Reactive thrombocytosis
- After bleeding, surgery, delivery
- IL-6 raises thrombopoietin
- Iron deficiency adds to it
- Settles as the cause resolves
Remember High platelets after PPH are usually reactive.
Exam trap Postpartum thrombocytosis is common and rarely needs a haematology work-up.
Why the other options are wrong
- Essential thrombocythaemia ET is a diagnosis of persistent, unexplained thrombocytosis.
- Chronic myeloid leukaemia CML would show high white cells and a left shift.
- Myelodysplasia with 5q deletion 5q− syndrome affects older adults with macrocytic anaemia.
- Splenic infarction after surgery No splenic event explains this picture.
A preterm baby with group B streptococcal sepsis has a neutrophil count that falls from 6 to 0.8 × 10⁹/L over 12 hours. Why do newborns with severe sepsis often become neutropenic rather than neutrophilic?
Answer: Small marrow neutrophil storage pool
A newborn's marrow holds a much smaller pool of mature neutrophils in storage than an adult's, and its precursor cells are already dividing near their maximum rate. In severe infection neutrophils leave the blood for tissues faster than the marrow can replace them, the storage pool is exhausted and the blood count falls. Neonatal neutrophils also show weaker chemotaxis and adhesion. Neutropenia in sepsis is therefore an ominous sign in newborns.
Neonatal neutrophil reserve
- Small storage pool
- Precursors near maximum output
- Sepsis exhausts reserves
- Neutropenia = poor sign
Remember Newborns run out of neutrophils quickly.
Exam trap A falling neutrophil count in a septic newborn is a warning, not improvement.
Why the other options are wrong
- Neonatal neutrophils live for months Neutrophils live hours to days, not months.
- Placental G-CSF stops after birth Neonatal G-CSF is made by the baby's own tissues.
- Lymphocytes suppress neonatal marrow Lymphocytes do not suppress the marrow in sepsis.
- Toxins destroy neutrophils in the blood Direct toxin destruction is not the main mechanism.
Where this sits in the CPSP syllabus
- Obstetrics & Gynaecology, Paper 2: Blood disorders in pregnancy and the perinatal period, White cell and platelet changes in the perinatal period; gestational thrombocytopenia
Read next
Drug dose pages
Doses, cautions and practical tips for drugs in this chapter.