Leukocytosis Differential Interpreter
Leukocytosis Differential Interpreter
A raised white cell count is not a diagnosis. The height of the count narrows the field a little, but it is the differential, the film and — when they point that way — the flow cytometry and the molecular test that decide between a reaction, a leukaemoid reaction and a leukaemia. This page works through the six features that actually discriminate and tells you which test settles it. It will not tell you the count is benign, because no count can.
Is this raised white count reactive, leukaemoid, or leukaemia?
Count + film pattern → where to look nextA 54-year-old presents with fatigue and early satiety. The white cell count is 62 × 10⁹/L with a neutrophilia, the film shows a continuous left shift through metamyelocytes to myelocytes and occasional promyelocytes with more myelocytes than metamyelocytes, the basophil count is 0.9 × 10⁹/L, no blasts are seen on a 200-cell differential, and the spleen is palpable 6 cm below the costal margin.
The six features that actually discriminate
2. Which lineage? Read absolute counts, never percentages.
3. How far back does the left shift go? Bands and metamyelocytes are a marrow response; a continuous shift to promyelocytes with a myelocyte peak is CML.
4. Are there blasts, and how many? Any blast needs explaining; 20% meets the acute leukaemia threshold in both classifications.
5. Are the basophils up? Reactions do not raise basophils. A myeloproliferative neoplasm does.
6. Is the spleen or are the nodes enlarged? Neither diagnoses nor excludes anything, but both change the urgency and the work-up.
Worked example
A 54-year-old presents with fatigue and early satiety. The white cell count is 62 × 10⁹/L with a neutrophilia, the film shows a continuous left shift through metamyelocytes to myelocytes and occasional promyelocytes with more myelocytes than metamyelocytes, the basophil count is 0.9 × 10⁹/L, no blasts are seen on a 200-cell differential, and the spleen is palpable 6 cm below the costal margin.
No blasts on the film, so neither blast rule applies
62 × 10⁹/L is high but not above 100, so this is not hyperleukocytosis and leukostasis is not the immediate question
Neutrophil lineage, a full left shift with a myelocyte peak, and an absolute basophilia of 0.9 → all three CML features are present together
The count alone would not have got here. At 62 × 10⁹/L this sits inside the leukaemoid range as well, and a leukaemoid reaction is the differential — it is the basophilia and the myelocyte peak that move it
Splenomegaly supports the diagnosis; its absence would not have excluded it
Action: BCR::ABL1 by RT-PCR on blood, with cytogenetics and FISH to follow. Not a leukocyte alkaline phosphatase score, which is obsolete and which the current ELN laboratory recommendations do not mention
Change one entry and the page changes with it. Drop the basophils to 0.05 × 10⁹/L and the same 62 × 10⁹/L with the same left shift reads as a leukaemoid reaction that cannot be separated from CML on the count — and still earns a BCR::ABL1
What the height of the count is and is not worth
| White cell count (× 10⁹/L) | What it narrows | What it does not settle |
|---|---|---|
| 11 to 30 | Leukocytosis is present | Nothing else. This range is overwhelmingly reactive and also contains early CML |
| 30 to 50 | Reactive causes are thinning out | A clonal disorder is still not established |
| 50 to 100 | The leukaemoid range | Nothing. Chronic phase CML has a median count of about 80 — the ranges overlap almost completely |
| Above 100 | Almost always a leukaemia or a myeloproliferative neoplasm | Which one, and whether leukostasis is present — cell type decides that, not the count |
Leukaemoid reaction against chronic myeloid leukaemia
| Feature | Leukaemoid reaction | Chronic myeloid leukaemia |
|---|---|---|
| Count | Usually 50 to 100 × 10⁹/L | Wide range, median about 80 × 10⁹/L |
| Left shift | Bands and metamyelocytes, tailing off | Continuous to promyelocytes, with a myelocyte peak |
| Basophils | Not raised | Raised in essentially every chronic phase case |
| Eosinophils | Variable | Raised in most cases |
| Neutrophil morphology | Toxic granulation, Döhle bodies, vacuolation | Usually unremarkable |
| Spleen | Not enlarged by the reaction itself | Commonly enlarged |
| Leukocyte alkaline phosphatase | High | Low |
| The test that decides | — | BCR::ABL1 by RT-PCR, with cytogenetics and FISH |
Why the film and the flow cytometry decide this, and the number does not
A white cell count is a single number standing in for several independent populations, each with its own kinetics. Neutrophils spend hours in the circulation, lymphocytes recirculate for years, and roughly half the neutrophils in the blood at any moment are not counted at all because they are marginated along the vessel walls. A dose of adrenaline or a flight of stairs releases that marginated pool within minutes and doubles the count without a single extra cell having been made. That is why an isolated high count taken during an acute event is worth very little, and why the shape of the differential carries far more information than the total.
The three things this page is asked to separate fail to separate on the number. A reactive neutrophilia can reach 50 × 10⁹/L in a severe infection. A leukaemoid reaction is conventionally defined above that figure and usually sits between 50 and 100. Chronic phase chronic myeloid leukaemia has a median count of around 80 and a range that runs from the barely abnormal to the enormous. Any threshold drawn through that overlap would misclassify in both directions, and the historical attempt to rescue the distinction with a leukocyte alkaline phosphatase score — low in CML, high in a reaction — has been abandoned because a molecular test answers the question definitively and a semi-quantitative cytochemical stain does not.
What does discriminate is structure. A reaction accelerates a normal process, so it releases cells that are one or two steps early and leaves the proportions between lineages broadly intact; it also marks the cells it releases, with toxic granulation, Döhle bodies and vacuolation. A clonal myeloid disorder expands the whole maturation sequence, so the film runs continuously back to promyelocytes and the myelocytes outnumber the metamyelocytes, producing a two-peaked distribution rather than a tail. And it expands lineages a reaction leaves alone: the absolute basophilia of chronic phase CML has no reactive counterpart worth the name. Blasts are the sharpest structural finding of all, because a blast in the blood of someone who is not recovering from marrow suppression represents a failure of maturation rather than an acceleration of it.
So the sequence is: read the absolute counts rather than the percentages, look at the film properly, and then send the test the film points to. In practice that means BCR::ABL1 by RT-PCR for a myeloid picture, flow cytometric immunophenotyping for a lymphoid one, and a marrow with cytogenetics and molecular studies for blasts. The count tells you how urgent that is. It does not tell you what it is.
Frequently asked questions
Can a white cell count on its own ever distinguish a leukaemoid reaction from chronic myeloid leukaemia?
No. A leukaemoid reaction is defined above 50 × 10⁹/L and usually runs between 50 and 100; chronic phase CML has a median count of about 80 × 10⁹/L and a range that covers the whole of that interval and beyond. The two distributions overlap almost completely, so no threshold separates them. The discriminators are the film — a continuous left shift with a myelocyte peak, and an absolute basophilia, in CML; toxic granulation, Döhle bodies and vacuolation in a reaction — and, definitively, BCR::ABL1 testing.
Should I request a leukocyte alkaline phosphatase score?
No. The score is low in CML and high in a leukaemoid reaction, which is why it was used, but it is a semi-quantitative cytochemical stain with poor reproducibility that has been replaced by BCR::ABL1 testing. The current European LeukemiaNet laboratory recommendations for the diagnosis and management of CML do not mention it anywhere, which says more about its status than any explicit statement of obsolescence would.
How high does the count have to be before it is dangerous in itself?
There is no accepted minimum. Hyperleukocytosis is defined above 100 × 10⁹/L, but leukostasis — impaired perfusion from the cell mass — depends far more on what the cells are than on how many there are. It is described in acute myeloid leukaemia at counts as low as 50 × 10⁹/L, because myeloblasts are large and poorly deformable, and it is rare in chronic lymphocytic leukaemia even at counts several times higher, because small mature lymphocytes pass through capillaries easily. Treat the symptoms — breathlessness, hypoxia, confusion, visual change, priapism — rather than the threshold.
Why does the page ask for absolute counts rather than percentages?
Because a percentage is a share of a total that is itself abnormal. If the neutrophil count triples, the lymphocyte percentage falls even though not one lymphocyte has been lost, and a reader working from percentages will see a lymphopenia that does not exist. Every threshold on this page — 7.0 × 10⁹/L for neutrophilia, 0.1 × 10⁹/L for basophilia, 0.5 × 10⁹/L for eosinophilia, 0.88 × 10⁹/L for monocytosis — is an absolute count. The absolute count calculators in this category convert a differential percentage into one.
The blast percentage is 12%. Is that acute leukaemia or not?
It depends on the genetics, and the two current classifications answer it differently. In the WHO 5th edition, a defining genetic abnormality such as PML::RARA, RUNX1::RUNX1T1, CBFB::MYH11 or mutated NPM1 makes it acute myeloid leukaemia at any blast percentage. In the ICC, those same entities require at least 10% blasts — which 12% meets — while a case with no defining abnormality and 10 to 19% blasts becomes MDS/AML rather than AML. Either way a blood blast percentage under 20 does not exclude acute leukaemia, and a marrow with cytogenetics and molecular studies is needed.
Related calculators
References
- Riley LK, Rupert J. Evaluation of patients with leukocytosis. Am Fam Physician. 2015;92(11):1004-1011.
- Sakka V, Tsiodras S, Giamarellos-Bourboulis EJ, Giamarellou H. An update on the etiology and diagnostic evaluation of a leukemoid reaction. Eur J Intern Med. 2006;17(6):394-398.
- Cross NCP, Ernst T, Branford S, et al. European LeukemiaNet laboratory recommendations for the diagnosis and management of chronic myeloid leukemia. Leukemia. 2023;37(11):2150-2167.
- American College of Emergency Physicians, Critical Care Medicine Section. Hyperleukocytosis and Leukostasis. 11 February 2025.
- Khoury JD, Solary E, Abla O, et al. The 5th edition of the World Health Organization classification of haematolymphoid tumours: myeloid and histiocytic/dendritic neoplasms. Leukemia. 2022;36(7):1703-1719.
- Arber DA, Orazi A, Hasserjian RP, et al. International Consensus Classification of myeloid neoplasms and acute leukemias: integrating morphologic, clinical, and genomic data. Blood. 2022;140(11):1200-1228.
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.