Pancytopenia Interpreter
Pancytopenia Interpreter
Marrow infiltration, megaloblastic anaemia, hypersplenism and aplasia can produce identical numbers. The counts tell you how urgent this is and nothing at all about why it is happening. Two things do: the reticulocyte response, which splits production failure from destruction and sequestration, and the blood film, which is the cheapest and most informative test in the whole work-up. This page grades each cytopenia by a named scale and then asks the two questions that actually narrow the differential.
All three lineages down — what is the mechanism?
Counts + retics + film → mechanismA 63-year-old man presents with six weeks of increasing fatigue and easy bruising. Haemoglobin 78 g/L, neutrophils 0.7 × 10⁹/L, platelets 32 × 10⁹/L. The reticulocyte count is 18 × 10⁹/L, which is low for the degree of anaemia. The film shows no dysplasia, no nucleated red cells, no blasts and no hypersegmentation. The spleen is not enlarged.
The reticulocyte count is the fork in the road
Reticulocytes raised → the marrow is producing and the cells are being lost. Immune destruction (Evans syndrome), consumption (thrombotic microangiopathy, disseminated intravascular coagulation) or sequestration (hypersplenism).
Read it as an absolute count or a production index. A reticulocyte percentage is a fraction of a red cell count that has itself fallen, so it rises when production has not.
- Pancytopenia
- Haemoglobin below about 120 g/L in women or 130 g/L in men, platelets below 150 × 10⁹/L, and white cells below 4.0 × 10⁹/L or neutrophils below about 1.8 × 10⁹/L
Worked example
A 63-year-old man presents with six weeks of increasing fatigue and easy bruising. Haemoglobin 78 g/L, neutrophils 0.7 × 10⁹/L, platelets 32 × 10⁹/L. The reticulocyte count is 18 × 10⁹/L, which is low for the degree of anaemia. The film shows no dysplasia, no nucleated red cells, no blasts and no hypersegmentation. The spleen is not enlarged.
Grade each cytopenia by CTCAE v5.0: haemoglobin 78 g/L is grade 3 anaemia (below 80); neutrophils 0.7 is grade 3 (1.0 down to 0.5); platelets 32 is grade 3 (50 down to 25)
Under CTCAE v6.0 the platelet grading changed — grade 3 now runs from 50 down to 10 and grade 4 begins below 10, so the same 32 is still grade 3 but a count of 15 would be grade 3 in v6.0 and grade 4 in v5.0. Name the version
The reticulocyte count is low → production failure, not destruction or consumption. That single result removes haemolysis, Evans syndrome and thrombotic microangiopathy from the differential
The film is unremarkable → no leukoerythroblastic picture, no dysplasia, no megaloblastic clue, no blasts. Nothing narrows it further from the blood
Neutrophils 0.7 is not below 0.5 and platelets 32 is not below 20, so this does not meet two of the three Camitta blood criteria for severe aplastic anaemia
Spleen not enlarged → sequestration is not contributing
So: marrow aspirate and trephine, with cytogenetics, flow cytometry and a PNH screen — after B12, folate, copper, HIV, hepatitis serology, thyroid, liver and renal profiles and a full drug history
Change one entry and the page changes. Make the reticulocyte count raised and the same three numbers become an emergency of a different kind — a direct antiglobulin test and a search for red cell fragments, today
Grading each cytopenia — and how the scale moved in 2026
| Grade | Anaemia, haemoglobin (g/L) | Neutrophils (× 10⁹/L) | Platelets, CTCAE v5.0 (× 10⁹/L) | Platelets, CTCAE v6.0 (× 10⁹/L) |
|---|---|---|---|---|
| 1 | Below normal to 100 | Below normal to 1.5 | Below normal to 75 | Below normal to 75 |
| 2 | 100 down to 80 | 1.5 down to 1.0 | 75 down to 50 | 75 down to 50 |
| 3 | Below 80, transfusion indicated | 1.0 down to 0.5 | 50 down to 25 | 50 down to 10, transfusion indicated |
| 4 | Life-threatening | Below 0.5 | Below 25 | Below 10, life-threatening |
Four diagnoses that produce identical numbers
| Diagnosis | Reticulocytes | Film | Spleen | What confirms it |
|---|---|---|---|---|
| Marrow infiltration or fibrosis | Low | Leukoerythroblastic, tear-drop cells | Often enlarged | Trephine biopsy — an aspirate is often dry or unrepresentative |
| Megaloblastic anaemia | Low | Hypersegmented neutrophils, oval macrocytes | Normal | B12 and folate; methylmalonic acid if borderline. Reticulocyte rise within days of replacement |
| Hypersplenism | Normal or raised | Often unremarkable | Enlarged | Imaging and the cause of the splenomegaly. Cytopenias are usually modest |
| Aplastic anaemia | Low | Unremarkable | Normal | Trephine showing under 25% cellularity; Camitta criteria for severity; PNH clone screen |
Why the counts cannot tell you the mechanism
Pancytopenia is a description of a full blood count, not a diagnosis, and the count itself carries almost no mechanistic information. A haemoglobin of 78 g/L with neutrophils of 0.7 and platelets of 32 is compatible with aplastic anaemia, with carcinoma filling the marrow, with severe B12 deficiency, with an autoimmune process destroying all three lineages, and with a large spleen doing nothing more sinister than holding onto cells. Those five have different investigations, different urgencies and different treatments, and nothing in the three numbers separates them.
The reticulocyte count does most of that work, and it does it because it measures production directly rather than inferring it. If the marrow can make red cells it makes reticulocytes, so a reticulocytosis in the face of anaemia means erythropoiesis is intact and the problem is downstream — destruction, consumption or sequestration. A low or inappropriately normal reticulocyte count means production has failed, and the whole marrow-replacement and marrow-failure differential opens up. One trap is worth stating plainly: the result must be read as an absolute count or as a production index, because a reticulocyte percentage is a fraction of a red cell count that has itself collapsed, and it rises when nothing has improved.
The film does the rest. Four patterns each point at a specific diagnosis and none of them is visible in the numbers. A leukoerythroblastic picture — nucleated red cells alongside immature granulocytes — means the marrow architecture is disturbed, by infiltration or fibrosis, and directs you to a trephine rather than an aspirate. Hypersegmented neutrophils with oval macrocytes mean megaloblastic anaemia, which is reversible and is missed precisely because its counts mimic marrow failure so exactly. Dysplastic granulocytes point at a myelodysplastic syndrome, after the reversible mimics have been excluded. And blasts change the timescale from weeks to the same day.
What the counts do determine is urgency, and that is worth grading formally rather than by impression. CTCAE gives named grades for each lineage, and naming the version now matters: version 6.0, released in 2025, moved the platelet grade 3 and 4 boundary from 25 to 10 × 10⁹/L and renamed the term, and the neutrophil grades moved too: the old grade 1 band (below the lower limit of normal down to 1.5 × 10⁹/L) is no longer graded at all, 1.0 to 1.5 becomes grade 1 where it had been grade 2, and grade 4 tightens from below 0.5 to below 0.1 — so a grade quoted without its edition is ambiguous for two of the three lineages. Severe aplastic anaemia has its own scale — the Camitta criteria, which combine a marrow cellularity below a quarter of normal with at least two of three peripheral thresholds — and that is a different instrument doing a different job. Grading the cytopenias tells you how fast to move. The reticulocyte count and the film tell you where to go.
Frequently asked questions
What actually counts as pancytopenia?
All three lineages reduced together: conventionally a haemoglobin below about 120 g/L in women or 130 g/L in men, a platelet count below 150 × 10⁹/L, and a white cell count below 4.0 × 10⁹/L or a neutrophil count below about 1.8 × 10⁹/L. Those limits vary with age, sex and ancestry and are laboratory-specific. A borderline neutrophil count in a person of African or Middle Eastern ancestry may reflect the benign Duffy-null associated neutrophil count rather than a cytopenia at all.
Why is the reticulocyte count the first thing to look at?
Because it splits the differential roughly in half and costs almost nothing. A raised reticulocyte count means the marrow is producing and the cells are being destroyed, consumed or sequestered — immune haemolysis with thrombocytopenia, a thrombotic microangiopathy, hypersplenism. A low or inappropriately normal count means production has failed, which opens the marrow-replacement and marrow-failure differential. Ask for the absolute count or a production index, not a percentage: the percentage rises simply because the red cell count has fallen.
Can I diagnose the cause without a bone marrow?
Sometimes. Megaloblastic anaemia is confirmed on B12 and folate with a reticulocyte response to replacement; hypersplenism with modest cytopenias and an obvious cause for the spleen may need no marrow; a drug cause that reverses on withdrawal speaks for itself. But a pancytopenia with a low reticulocyte count, an unremarkable film and no reversible cause found needs an aspirate and trephine, and a leukoerythroblastic film needs a trephine specifically, because infiltration and fibrosis frequently give a dry or unrepresentative aspirate.
Which CTCAE version should I grade against?
Whichever your protocol or trial specifies, and say which one. Version 6.0 was released in 2025, with an errata in January 2026, and changed the platelet term: it is now called Thrombocytopenia, grade 3 runs from below 50 down to 10 × 10⁹/L with transfusion indicated, and grade 4 is below 10 with life-threatening consequences. Under version 5.0 the boundary was 25. A platelet count of 15 is therefore grade 4 in v5.0 and grade 3 in v6.0. The neutrophil grades moved as well — the neutrophil grades moved too: the old grade 1 band (below the lower limit of normal down to 1.5 × 10⁹/L) is no longer graded at all, 1.0 to 1.5 becomes grade 1 where it had been grade 2, and grade 4 tightens from below 0.5 to below 0.1 — so a neutrophil count of 0.3 × 10⁹/L is grade 4 on v5.0 and grade 3 on v6.0. Only anaemia is unchanged.
What are the Camitta criteria and how do they differ from CTCAE grading?
They do a different job. CTCAE grades the severity of each cytopenia separately, for adverse event reporting. The Camitta criteria define severe aplastic anaemia as a disease: a marrow of less than 25 per cent of normal cellularity together with at least two of a neutrophil count below 0.5 × 10⁹/L, a platelet count below 20 × 10⁹/L, and a reticulocyte count below 20 × 10⁹/L. Very severe disease is the same with a neutrophil count below 0.2 × 10⁹/L. The marrow cellularity is part of the definition, so the criteria cannot be applied from the blood count alone.
Related calculators
References
- Chiravuri S, De Jesus O. Pancytopenia. In: StatPearls. Treasure Island (FL): StatPearls Publishing; updated 23 August 2023.
- US Department of Health and Human Services, National Cancer Institute. Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. Published 27 November 2017.
- US Department of Health and Human Services, National Cancer Institute. Common Terminology Criteria for Adverse Events (CTCAE) version 6.0. Released 2025, errata January 2026.
- Camitta BM, Rappeport JM, Parkman R, Nathan DG. Selection of patients for bone marrow transplantation in severe aplastic anemia. Blood. 1975;45(3):355-363, as applied in the EWOG-SAA diagnostic work-up recommendations v1.0.
- US Department of Health and Human Services, National Cancer Institute. CTCAE version 6.0, released 2025 (errata January 2026), and the v5.0 to v6.0 mapping resource.
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.