Quick tip: A high Hb is often not due to a primary haematological problem.
An elevated Hb can be due to either a primary myeloproliferative neoplasm (e.g polycythaemia vera), or may be secondary to other conditions. Secondary causes are either due to hypoxia, or less commonly excess erythropoietin or endocrine disorders. It will often be identified as an incidental finding on a full blood count, but may present with thrombosis, more often arterial than venous, or constitutional symptoms of hyperviscosity such as headaches, itching, blurred vision parasthesiae or chest pain. Haematocrit is probably a more reliable measure of polycythaemia than haemoglobin alone-this is the percentage of blood made up by red cells i.e 0.45 = 45% of centrifuged blood is red blood cells.
Referral to haematology should be considered in patients with a persistently elevated haemoglobin in the absence of a respiratory cause (or if the Haematocrit is over 0.48 (48%) in females or 0.52 (52%) in males, discussion should be had with haematology).
If contacting a haematology on-call service before referring the patient consider asking if testing for JAK2 V617F mutation would be appropriate as this may improve efficiency of haematology outpatient review. This is an acquired mutation presented in around 95% of patients with polycythaemia vera.
An urgent referral should be sent in those presenting with thrombosis or symptoms of hyperviscosity for urgent venesection. Aspirin is recommended in primary polycythaemia but not secondary polycythaemia, after the appropriate risk assessment for risks and benefits of aspirin use.
Secondary polycythaemia can be managed by venesection in certain cases although evidence is often weak (see BSH guideline).
Links: BMJ NICE CKS BSH Guideline
Quick Tip; Mean corpuscular volume is an average reading of red cell size, therefore a 'normal' MCV could be a combination of a micro- and macrocytic process
Anaemia investigations involve determining whether it is an intrinsic problem with the bone marrow or due to an extrinsic cause. This process can be made easier by classifying the anaemia by red cell size;
Refer to haematology if concerned about underlying haematological or bone marrow problem such as a raised paraprotein which may indicate myeloma (see Paraprotein and Light chains sections), or myelodysplasia (macrocytosis in the absence of other obvious causes, especially if other cell lines affected e.g low platelets/ white blood cells)
Iron replacement strategies
Research into iron replacement strategies (currently in non-anaemic, iron deficient women but theory likely to be applicable to anaemic patients) shows that alternate day, once daily oral iron dosing optimises absorption and reduces side effects.
BD/TDS dosing regimens cause significant homeostatic hepcidin rises which actually inhibits subsequent iron absorption from the gut. The ‘spare’/unabsorbed iron is then responsible for many of the GI side effects. Alternate day dosing or once daily dosing allows time for hepcidin to drop again after first dose, and thus allow absorption of subsequent doses.
Although not definitive, this appears to be the future for oral iron replacement. Consider this if patients have previously been intolerant of traditional regimen oral iron replacement before referring for IV iron-much cheaper and less (& reversible) significant side-effects than IV iron (skin staining etc).
B12/folate replacement strategies (Adapted from NICE CKS B12 deficiency guidance)
Iron studies interpretation
Tests for a patient's iron status can be confusing. Those most commonly used are;
Iron absorption from oral iron supplements given on consecutive versus alternate days and as single morning doses versus twice-daily split dosing in iron-depleted women: two open-label, randomised controlled trials. Stoffel NU et al. Lancet Haematol. 2017 Nov;4(11):e524-e533. doi: 10.1016/S2352-3026(17)30182-5
Guideline for the laboratory diagnosis of iron deficiency in adults (excluding pregnancy) and children. Br J Haematol. A British Society for Haematology Good Practice Paper. Fletcher A, et al. 2022 Feb;196(3):523-529. doi: 10.1111/bjh.17900. Epub 2021 Oct 24. PMID: 34693519.
Links: BMJ B12 and Folate deficiency investigations
Quick Tip; If a patient is B12 and folate deficient, B12 MUST be replaced before folate as the surge in red cell production if folate is given first can causes a dramatic consumption of remaining B12 and risk of subacute combined degeneration of the cord
An increased mean corpuscular volume is seen due to abnormal DNA production, conditions which increase the proportion of immature cells in the circulation, or when red cells are exposed to toxic substances.
It is possible to get falsely 'normal' B12 assays even with B12 deficiency – if strong clinical suspicion, consider discussion with biochemistry lab/haematology as other investigations may be appropriate.
Referral to haematology is recommended if haemolysis, dysplasia or myeloma are identified on the bloods and blood film, or if other cytopenias are present (thrombocytopenia/ leukopenia).
Links: BMJ Best Practice (subscription required) NICE CKS B12 and Folate deficiency investigations BSH B12/folate deficiency guidance
Quick tip; Microcytic anaemia in the UK is most commonly caused by iron deficiency
The differential for microcytic anaemia is quite limited. The most important causes are;
If no evidence of thalassaemia, referral to gastroenterology should be considered because primary haematological causes for microcytosis are so rare. Identification of a source of malabsorption or bleeding is the priority. If associated with changes in bowel habit or an older patient particularly, consider 2 week wait referral.
Iron replacement strategies
Research into iron replacement strategies (only currently in non-anaemic, iron deficient women so far but likely to be applicable to anaemic patients) shows that alternate day, once daily oral iron dosing allows time for hepcidin to drop again after first dose (it rises after iron absorbed). Daily dosing, and especially split dosing causes hepcidin rises which blocks further absorption from the gut. The ‘spare’/unabsorbed iron is then responsible for many of the GI side effects. Although not definitive, this appears to be the future for oral iron replacement. Consider this if patients have previously been intolerant of traditional regimen oral iron replacement before referring for IV iron-much cheaper and less (& reversible) significant side-effects than IV iron (skin staining etc).
Links; NICE CKS
Quick Tip; Although a B12 level above the reference range has an association with cancer, there is no guidance about a suggested investigative pathways
B12 levels above the reference range are not an uncommon finding in those who are not on B12 supplementation. A Danish population cohort study (n=333,667) found that the incidence of B12 levels above the reference range was 6%.
A raised B12 level (without supplementation) has been associated with increased cancer incidence. An increasing level of B12 seems to correlate with increasing cancer incidence but this appears to be lost from 1 year after the initial measurement. The 3 commonest cancers were myeloid haematological malignancy, liver malignancy and myeloma (although the myeloma association was not seen in another study (Jammal et al.)). Many of the commonest solid organ cancers represented the remaining associated malignancies.
200-600pmol/L (reference range)
Standardised cancer incidence ratio with different baseline B12 levels. Adapted from Arendt JF, Pedersen L et al (2013).
Cancer however is not the only cause of raised B12, with other causes including liver disease, alcoholism, interstitial renal disease, autoimmune and infectious disease (Arendt, Jammal).
Due to the relatively low cancer incidence with even high B12 levels it is hard to create an investigative/management pathway. If an incidental raised B12 is found, it seems reasonable to undertake the following;
Quick Tip; An acutely raised high platelet count is most commonly reactive due to infection or bleeding and often needs only observation
Thrombocytosis is a common finding in hospital inpatients as a reactive feature to acute illness. Persistent or significantly raised platelet counts (>450 x 109/L), especially in the absence of an inflammatory trigger, should be referred for investigation for a possible underlying myeloproliferative disorder (essential thrombocythaemia, polycythaemia vera or myelofibrosis).
Workup of chronic thrombocytosis
If on repeat an elevated count is persistent and no other trigger is obvious, refer to haematology.
If significantly elevated (>600 x 109/L) or the patient has an increased risk of thrombosis (history of acute coronary syndrome, venous thromboembolism or cerebrovascular disease) or is aged over 60, consider referring more urgently to haematology services/ seek on-call haematology advice as these features all increase the thrombosis risk
Essential thrombocythemia/myeloproliferative disorder
Links: BMJ (subscription required) BSH Guideline
Quick Tip; Bleeding from mucosal surfaces such as the gums or tongue in the context of thrombocytopenia are associated with an increased risk of serious bleeds such as intracranial bleeds therefore require urgent assessment & management
The common causes for thrombocytopenia can be broadly divided into more acute or chronic causes.
* Some of the medications most frequently associated with thrombocytopenia include; Heparin, quinine, penicillins, cephalosporins, rifampicin, NSAIDS, anticonvulsants, disease-modifying anti-rheumatic drugs, tirofiban, amiodarone, furosemide (See refs)
ITP does not have any specific diagnostic tests therefore referral to haematology should be performed in the case of a prolonged or severe thrombocytopenia, and when non-haematological causes seem unlikely.
See "Transfusion Threshold" module for advice on safe platelet levels peri-operatively and pre-procedures
Visentin G, Liu C. Drug-Induced Thrombocytopenia. Hematology/Oncology Clinics of North America. Volume 21, Issue 4, Pages 685-696
Aster R, et al. Drug-induced immune thrombocytopenia: pathogenesis, diagnosis, and management. Journal of Thrombosis and Haemostasis. Volume 7, Issue 6, June 2009, Pages 911–918
Quick Tip; Unless significantly raised the white cell count itself does not indicate likelihood of an underlying haematological disorder, as leukaemia and lymphoma may have a normal or low white cell count therefore a thorough assessment with blood film is critical
Broadly speaking a high white cell count will be either due to increased myeloid cells (neutrophils or less commonly eosinophils, basophils, blasts) or lymphocytes. The pathology also varies hugely according to the rate of change and the type of cell. See below for a simplified strategy to think about leukocytosis;
|Cell type\Speed of change||Days/weeks||Months/years|
|Mature cells (neutrophils/ lymphocytes)||Acute infection/ inflammation|| |
Chronic leukaemia, myeloproliferative neoplasms, chronic infections
|Immature cells (blasts aka leukaemia cells)||Acute leukaemia||Not seen (as cause of isolated raised white cell count)|
Commonest acute causes (days/weeks) of leukocytosis;
Commonest chronic causes (months/years) of leukocytosis;
Investigations are merited if;
If the blood film reports that a malignancy such as leukaemia is likely, haematology will be involved by the lab team, but you can see the 'Workups' submenu for next steps in investigation.
Quick Tip; Neutrophilia is only very rarely is caused by a malignant problem
Almost all cases of neutrophilia are due to an acute insult, and reactive counts of over 50x109/L can sometimes be seen
Review of trend is the most important feature in assessing neutrophilia.
These are some features to help identify probable aetiology
Bacterial infection- Acute rise, will drop quickly after insult resolves. Can sometimes rise even above 50x109/l- no specific action needed unless the neutrophilia fails to resolve within a few short weeks after the insult has
G-CSF use in patients on chemotherapy – can also give an acute rise with very high neutrophils, but will also resolve as the G-CSF is stopped.
Smoking can lead to a persistent neutrophilia (see reference below about blood donors- neutrophilia seen in 22% of smokers and 2% of non-smokers)
CMML is a slow-growing leukaemia which usually presents as an incidental finding of chronic, slowly rising neutrophilia and monocytosis in a middle aged/older person. This comes on over months/years. The blood film will usually show dysplastic features. It would require a non-urgent referral to haematology (although officially is a malignancy so see local protocols for referral guidance). If other counts (haemoglobin and platelets) had reduced, or there were systemic symptoms e.g splenomegaly/weight loss, urgent referral is recommended.
Chronic myeloid leukaemia is also slow-growing and usually associated with basophilia (fairly unique to CML and other myeloproliferative neoplasms), thrombocytosis and splenomegaly. This would also merit a non-urgent referral to haematology so long as no blasts are seen on the blood film and there are no other cytopenias (although officially is a malignancy so see local protocols for referral guidance).
Chronic neutrophilic leukaemia is very rare and associated with an isolated, persistent and progressive neutrophilia.
Smoking as a Cause for Mild Chronic Neutrophilia. Stemmelin G et al. Blood 2004 104:3796.
Links: Elsevier summary
Quick Tip; A count of 1 x 109/L or more is unlikely to be associated with a significant risk of infection
Most cases of incidental neutropenia will resolve on a repeat after a reasonable period as they will be due to transient infection. For those that do not resolve, around a third will be chronic idiopathic (see ref Lima et al.-note small study population), with other causes listed as below.
Neutropenia is defined as mild: 1-1.5 x109/L, moderate: 0.5-1 x109/L or severe: <0.5 x109/L.
If neutrophil count >1 x109/L, chronic and the patient has no clinical features of neutropenia (frequent infections, mouth ulcers) and initial investigations for an underlying disorder are negative then it is likely that no further action is needed apart from a repeat count in a few months to measure the trend in neutrophil count. If any clinical concern, especially in the context of other low cell counts or evidence of infection/organomegaly/mouth ulcers/fever then discuss with haematology.
In the case of fever and neutropenia prompt treatment with antibiotics is critical, especially in those patients with a neutrophil count of less than 1 x109/L.
Links: BMJ (needs subscription) BMJ best practice (needs subscription)
Quick tip - A normal lymphocyte count does not rule out lymphoma
Lymphocytosis will often be identified as an incidental finding on routine bloods. Many cases will be benign and some may indicate chronic haematological malignancy. Acute leukaemia will not be reported on a full blood count as a lymphocytosis, because the cells 'look' different to the analyser machines. In the case of a leukaemia, the analyser 'flags' the cells as abnormal and the laboratory performing the test will urgently review the blood film to see if it is leukaemia.
Referral to haematology is recommended in the case of persistent lymphocytosis, especially when associated with B symptoms, lymphadenopathy or hepatosplenomegaly.
A persistent lymphocytosis under 10x109/L in an otherwise well patient may be reported as 'Monoclonal B cell lymphocytosis' on the blood film. This condition is a CLL precursor which needs monitoring alone, as without any clinical features it needs no treatment. Once the lymphocyte count is over 10x109/L it is more likely to have a blood film comment of 'suspected CLL' possibly with advice for referral.
Caring for known CLL in the community
The CLL disease course is most often slowly progressive and a slow increase in the lymphocyte count is expected. Features that would require a more urgent referral to haematology or more prompt follow up than planned would be a rapidly rising white cell count (doubling time less than 6 months), any new cytopenias, progressive lymphadenopathy, and progressive constitutional symptoms such as fevers or sweats.
Links: BMJ (needs subscription)
Quick tip; HIV is the most significant cause of lymphopenia, and otherwise often has little clinical significance
Lymphopenia is almost always due to reversible causes and a count of over 0.5 x 109/L is rarely of clinical significance.
If lymphopenia is mild (>1 x109/L) and remains stable over a 6-12 month period with no other indicators of underlying disease then no specialty referral is needed as there would be no treatment for it and it does not confer a clinical risk. Persistent lymphopenia of <1 x109/L would merit further investigation as above and referral to the appropriate medical specialty
Lymphopenia is incredibly common in clinical practice and haematological causes for it are rare therefore referral to haematology is only recommended if there is other evidence to support underlying haematological problems like other cytopenias (e.g suspected MDS) or symptoms or signs in keeping with lymphoma (weight loss, fevers, sweats, lymphadenopathy)
Links: BMJ (requires subscription)
Quick Tip; Eosinophilia has a huge variety of causes, and a thorough systems review is required to plan investigations
Eosinophilia is a relatively uncommon finding but elevated counts should be investigated if persistent or there is associated clinical concern. Eosinophils are involved in fighting parasites and in atopy and allergy.
A history should concentrate on;
According to BSH guidance if there is a count of 0.5-1.5 x 109 and no end organ damage and a probable cause from the history then no further investigations may be required. If however, there is a probable cause that merits investigation, end organ damage or eosinophilia persistently over 1.5 x 109 then further investigation is merited. This would traditionally be with haematology unless there is a recent history of foreign travel and suspicion of helminthic infection in which case an infectious disease referral may be more appropriate.
Links: BSH Guideline BMJ (requires subscription)
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4515759/ (Eosinophilia end-organ damage)
Quick Tip; Chronic monocytosis is most commonly caused by chronic myelomonocytic leukaemia in older patients
Monocytes are a phagocytic cell from the myeloid lineage (which includes neutrophils, platelets, eosinophils etc). Monocytosis is most often an acute finding, and when chronic has a relatively limited differential diagnosis.
Acute monocytosis common causes;
Chronic monocytosis common causes;
Chronic myelomonocytic leukaemia
CMML is a slowly progressive leukaemia which usually presents as an incidental finding of chronic, slowly rising monocytosis +- neutrophilia in an older person. The blood film will probably show dysplastic features. It would require a non-urgent referral to haematology. A caveat to this is that if other counts (haemoglobin, neutrophils or platelets) had reduced, or there were systemic symptoms e.g splenomegaly/weight loss, then urgent referral would be recommended.
If there are results dating back over a long period of a stable or very slowly growing monocyte count (usually remains under 10), with normal haemoglobin and platelets and no systemic symptoms (e.g hepatosplenomegaly), this could pragmatically continue to be observed with a full blood count at reasonable intervals. Many older patients will die from other causes before the CMML ever becomes a problem.
If there is a patient who is known to have CMML and they present with acute anaemia, thrombocytopenia or neutropenia and they appear unwell, this may be acute leukaemia as CMML can transform to AML-seek urgent haematology opinion if this is suspected.
Lynch, DT, Hall, J, Foucar, K. How I investigate monocytosis. Int J Lab Hem. 2018; 40: 107– 114. https://doi.org/10.1111/ijlh.12776
Quick tip - 90% of patients with a ferritin level 300-1000 µg/L in routine medical practice will not have an iron overload disorder
Raised ferritin is incredibly common in clinical practice, but the important feature to consider is whether there is any iron loading into organs such as the liver. 90% of patients with a raised ferritin in routine medical practice will not have an iron loading disorder [Olynyk], and instead their raised ferritin will be secondary to inflammation or another organ disease-liver disease for example.
Even at levels of over 1500µg/L, cancer, inflammation, liver disease and alcohol abuse were all still more common than an iron loading disorder [Hearnshaw].
Causes of raised ferritin (e.g. >300 µg/L in men and postmenopausal women; >200 µg/L in premenopausal women-check local reference ranges)
HFE genotyping aims to identify the two commonest mutations seen in the HFE gene, and thus in hereditary haemochromatosis; C282Y and H63D. Homozygotes for C282Y represent the majority of those with genetic haemochromatosis in the UK. Heterozygous C282Y is the carrier state, and not associated with iron loading when the sole mutation. The H63D mutation appears to carry a much lower risk of iron loading. A compound heterozygote of C282Y and H63D mutations (1 of each allele) can also lead to an iron loading phenotype but usually only if there are other iron-loading pathologies.
Olynyk JK, Cullen DJ, Aquila S, et al. Population based study of the clinical expression of the haemochromatosis gene. N Engl J Med1999;341:718-24.
Hearnshaw, S., Thompson, N.P. & McGill, A. (2006) The epidemiology of hyperferritinaemia. World Journal of Gastroenterology, 12, 5866–5869.
Links: BMJ (subscription required) GP journal summary article BSH Hyperferritinaemia guideline
Quick tip; When a paraprotein has been identified, the most important step is to identify whether end organ damage from any possible underlying disease is present (bone pain, anaemia, new/progressive renal impairment or signs of lymphoma)
A paraprotein may be found as part of a workup for anaemia of unclear cause or during investigation of pathological fracture, and occasionally as an incidental finding. Below are the common causes and recommended further management of these patients. There is a supplemental 'notes' section below giving further information on paraproteins and their significance.
If identified, a paraprotein will be reported as the paraprotein heavy chain (in black in the image below); IgG, IgA, IgM, IgD, IgE (listed in decreasing frequency) and the associated light chain (in purple in the image below); kappa or lambda, e.g IgG kappa of 7g/L
Another description of the paraprotein may be as a 'band' in the 'gamma/beta zone' - this is the area on the immunofixation that corresponds to paraproteins. Other zones such as alpha and sometimes beta correspond to inflammation or other autoimmune/inflammatory disorders rather than myeloma/MGUS. Polyclonal rises are seen in inflammation and are not indicative of myeloma/MGUS. Again, a significant result such as a paraprotein should be specifically reported-if unclear, call the immunology lab for clarification of the result.
If there is no paraprotein detected, and the serum free light chain results are normal, the probability if myeloma is incredibly small, as only 3-5% of patients have a myeloma which produces no clonal protein ('non-secretory' myeloma).
Links: BMJ (requires subscription) BMJ Best Practice (requires subscription)
Quick tip; The most relevant part of a serum free light chain test is an abnormal ratio of kappa to lambda rather than increases in the kappa or lambda individually
Light chains are a distal fragment of the immunoglobulin protein (purple lines on the below image)
In myeloma and other plasma cell disorders, the malignant plasma cell makes the same protein over and over again.
75% of myeloma cases secrete a protein with both a heavy (e.g IgG, IgA) and light chain (kappa or lambda). 20% secrete light chains only. 3-5% produce no protein at all ('non-secretory' myeloma).
The light chain can only be kappa or lambda, and therefore if one is produced clonally (again and again by malignant cell), it will increase significantly in proportion to the other.
A standard serum free light chain screen will report Kappa and Lambda light chain levels and a ratio of Kappa:Lamba. The reference range varies between hospitals- please see your local reference range.
A proportional rise in light chain subtypes is seen in inflammation and renal impairment. For example a patient has both kappa and lambda levels raised to around 150mg/L, but the ratio of one to the other is maintained. Alternatively there may be one that is only just outside the reference range.
A significant imbalance in this ratio however may be in keeping with myeloma; either as part of a paraprotein or as part of a light-chain only myeloma.
The above results are reference ranges are purely for educational purposes and each set of results must be interpreted in the clinical context of the presentation.
Note that the inflammation example has a ratio outside of the reference range, but levels of the kappa light chain are only mildly increased. In this example, it would usually be advised to repeat the test when the inflammatory episode has resolved.
Note in renal impairment, a wider ‘normal’ range for the light chain ratio is used (impaired kidneys cannot clear light chains as effectively). Rather than 0.26-1.65, a range more like 0.4-3.1 (for example-check local range) is used. Therefore even if the light chain amounts are increased, even up to almost 200 for example, it is still most likely due to the renal impairment so long as the ratio is in range.
If a significant discrepancy is identified between the light chains, as in the bottom example, investigate as per the Myeloma workup section.