Need to Know - Haematology
Haematologists Dr Rob Weinkove and Dr Beverley Hunt answer GP Dr Colin Kenny’s questions on polycythaemia, sickle cell, thallasaemias and thrombophilia
Haematologists Dr Rob Weinkove and Dr Beverley Hunt answer GP Dr Colin Kenny's questions on polycythaemia, sickle cell, thallasaemias and thrombophilia
Polycythaemia or erythrocytosis is an increase in red cell mass. The most useful way to detect this is using the haematocrit – often referred to as Hct – or packed cell volume (PCV), which is reported as a percentage or as a proportion. A haematocrit of >0.51 (or 51%) for a man or >0.48 (or 48%) for a woman is raised.
Polycythaemia can be either true – a genuine increase in red cell mass – or apparent – a normal red cell mass but with a reduction in plasma volume.
Apparent polycythaemia is seen with diuretics, after alcohol or transiently in dehydration.
True polycythaemia is either caused by polycythaemia vera – an uncontrolled proliferative bone marrow disease – or secondary to chronic hypoxia from smoking, chronic lung disease, cyanotic heart disease or obstructive sleep apnoea. Rarely it can be caused by malignancies producing erythropoietin such as renal cell, hepatocellular or lung cancer.
If a patient has polycythaemia, repeat the full blood count if you suspect dehydration. Otherwise, the patient could be advised to stop smoking and drinking alcohol, and diuretics should be reduced or stopped if feasible.
Guidelines for the management of polycythaemia are available at www.bcshguidelines.com.
Patients with apparent polycythaemia – or with a mild polycythaemia secondary to smoking – could donate blood provided there are no other contraindications.
What differences in reference ranges should we be aware of in different ethnic groups, for example neutropenia in Afro-Caribbeans?
Reference ranges for blood tests are determined within each laboratory after testing a number of healthy individuals locally.
Most reference ranges encompass 95% of results, so a healthy person would have a 5% chance of lying outside the ‘normal' range. Thus reference ranges will not take account of differing blood count parameters at the extremes of age, in pregnant women or some ethnic groups.
Patients of Afro-Caribbean origin have lower neutrophil counts – and consequently lower total white cell counts – than other ethnic groups. This is physiological, and does not produce a higher infection rate.
One study quoted a neutrophil reference range for Afro-Caribbean men of 1.0 to 5.8x109/l, compared with 1.7 to 6.1 for white men. Mild neutropenia in the range 0.9 to 1.5 x109/l can be normal in a healthy Afro-Caribbean person.
A reasonable strategy for an Afro-Caribbean patient with isolated mild neutropenia is to repeat the blood count in a few weeks. If the neutrophil count is stable and the patient is otherwise healthy, there should be no concern.
Indications for referral or for discussion with a haematologist include recurrent infections, additional unexplained blood count abnormalities or severe neutropenia – that is neutrophils below 0.5.
Platelet counts can also be lower in Afro-Caribbeans and platelet counts in the range 100-150 x109/l can be normal.
No further investigation is needed if there are no symptoms of bleeding, no other significant blood count abnormalities and the platelet count is stable.
Indications for referral or discussion with a haematologist include bleeding symptoms, additional blood count abnormalities or persistent moderate thrombocytopenia (platelets <100 x109/l).
And what about differences as a result of pregnancy?
The blood count changes in pregnancy include a dilutional fall in haemoglobin as the rise in red cell mass is outstripped by a greater rise plasma volume. Haemoglobin levels as low as 10 g/dl can be normal in the third trimester. However, both folic acid and iron deficiency are prevalent in pregnancy, so exclude these if the haemoglobin falls below 10.5 g/dL.
Mild neutrophilia -up to 13 x109/l) is common in late pregnancy. Platelet counts usually remain stable but some women develop gestational thrombocytopenia, with a gradual fall in platelet count towards the end of pregnancy and spontaneous resolution afterwards. Bleeding complications are rare.
Pre-eclampsia and HELLP syndrome must be excluded in pregnant women with thrombocytopenia. Platelet counts of <80 x109/l should prompt urgent referral, as this level could preclude epidural anaesthesia and thus intervention may be warranted.
How should a patient with anaemia and a raised MCV be investigated? How should patients on vitamin B12 injections be monitored?
Anaemia with a raised mean corpuscular volume (MCV) is known as macrocytic anaemia. Causes include alcohol, vitamin B12 or folate deficiency, hypothyroidism, liver disease and rarely haemolytic anaemia or myelodysplasia.
A clinical and dietary history may point to the likely cause. It is particularly worth asking about any personal or family history of autoimmune disease, and asking if the patient is vegan or vegetarian.
The first-line investigations should include vitamin B12 level, folate level, thyroid stimulating hormone, liver profile, reticulocyte count and a blood film.
Vitamin B12 is only present in meat and dairy products so the causes of low or borderline low vitamin B12 include dietary B12 deficiency in vegans or in those with poor diet.
B12 malabsorption can be due to pernicious anaemia (PA), gastrointestinal disease or surgery affecting the stomach or distal ileum.
The Schilling test was used to indicate the cause of vitamin B12 deficiency but is no longer generally available. Certainly intrinsic factor antibodies are useful in confirming PA but are not positive in all cases.
Management is usually based on the clinical history. If there is a history of poor diet, no evidence of B12 malabsorption (no gastrointestinal surgery, negative intrinsic factor and parietal cell antibodies) and no evidence of neuropathy it is reasonable to give a trial of oral vitamin B12 replacement for example cyanocobalamin tablets 50µg OD.
For patients on oral replacement, the vitamin B12 level should be rechecked after two to three months and – if still low – presume probable malabsorption and start parenteral B12.
For patients with the suggestion of B12 malabsorption or with neuropathy, parenteral vitamin B12 replacement must be used. The initial replacement consists of six injections of 1mg intramuscular hydroxobalamin over two or three weeks. There is no need to recheck the vitamin B12 level.
After the initial injections, hydroxocobalamin should be continued at a dose of 1mg IM every two to three months for life. In PA it is not good practice or cost-effective to wait for the vitamin B12 levels to fall before the next injection.
What are the pitfalls in making a diagnosis of multiple myeloma?
Because myeloma can cause bone disease and can produce a paraprotein with variable properties, the range of presenting symptoms is broad.
A ‘myeloma screen' consists of full blood count, ESR or plasma viscosity (whichever is available), serum and urine electrophoresis, renal function tests and bone profile. In addition, X-ray of any symptomatic area (looking for fracture, osteopenia or lytic lesion) should be considered in patients over 40 years with any of the following: unexplained early osteoporosis (particularly unprovoked vertebral compression fracture), unexplained normocytic anaemia with high ESR or plasma viscosity (after other causes of anaemia are excluded), unexplained hypercalcaemia, and new onset of Raynaud's phenomenon. Other clinical features might include renal impairment or recurrent bacterial infections but screening all such patients for myeloma would yield a very low pick-up rate.
In 97% of cases, myeloma produces a paraprotein that can be detected by electrophoresis of serum or urine. It is essential to send both serum and urine, as in a third of cases a paraprotein will be detectable in urine but not serum.
Electrophoresis has a very high false positive rate because of the prevalence of benign paraproteins. The incidence of ‘monoclonal gammopathy of uncertain significance' (MGUS) rises with age, reaching 5% in those over 70 years. Some other haematological conditions can produce paraproteins, such as non-Hodgkins lymphomas, chronic lymphocytic leukaemia and amyloidosis.
Distinguishing myeloma from MGUS relies on finding a monoclonal plasma cell population on biopsy (of bone marrow or of a plasmacytoma), and demonstrating that end-organ damage has occurred (such as renal failure, bony lytic lesions, hypercalcaemia or anaemia). Absence of end-organ damage and a paraprotein level of <10g/L is typical of MGUS.
Occasional patients with MGUS may go on to develop myeloma so they do need follow-up. MGUS should be monitored six-monthly or annually to look for an increase in paraprotein level or the development of bony symptoms, hypercalcaemia, renal failure or anaemia.
Haemoglobinopathies are often confusing as there are so many combinations. What are the most important points about the commonest combinations, and is there a good online resource where I can refresh my memory?
Haemoglobinopathies are grouped into two broad categories: those where an abnormal haemoglobin chain is produced including sickle cell disease or haemoglobin C disease and those in which reduced amounts of a haemoglobin chain are made – the thalassaemias.
The most common and clinically significant haemoglobinopathies to be aware of are sickle cell disease and thalassaemia.
Sickle haemoglobin is prevalent in sub-Saharan Africa, the Middle East and India. A point mutation in the ß-haemoglobin gene predisposes red cells to sickling in conditions of hypoxia or acidosis. Sickle cell disease is recessive and if just one copy of the abnormal gene is inherited, the patient is said to have sickle trait.
Sickle trait is asymptomatic, and does not normally affect life expectancy. Some patients who coinherit one copy of the sickle gene with a second ß-haemoglobin gene abnormality have a sickling disorder similar to homozygous sickle cell disease – for example sickle cell disease or Sß-thalassaemia disease.
Patients with sickle cell disease are prone to a range of complications. Children may develop painful dactylitis due to red cell sickling and vaso-occlusion in the digits. Splenic infarction is universal, and in children can acutely lead to a ‘sequestration crisis' where red cells are sequestered in a rapidly enlarging spleen. Later in life, all patients become hyposplenic and require appropriate vaccination and antibiotic prophylaxis.
There is an increased red cell turnover in sickle cell disease, so conditions that reduce red cell production, such as parvovirus infection or folate deficiency, can lead to an aplastic crisis. The most common crisis, however, is the painful vaso-occlusive crisis, which can occur at times of cold, stress, hypoxia or infection. These are treated with analgesia, fluids and keeping warm.
Other long-term complications of sickle cell disease include stroke, avascular necrosis of the femoral or humoral head, retinopathy, nephropathy, pulmonary hypertension, priapism and erectile dysfunction. All patients with sickle cell disease should be referred for screening and pre-emptive management of complications. Click here for a clinical review of sickle cell disease from the BMJ.
Thalassaemias are conditions where one or more of the haemoglobin chains (either ? or ß) are produced in a reduced quantity. They are a very heterogeneous group of conditions, as there are hundreds of causative mutations. Because the haemoglobin chains produced in thalassaemia are normal, standard haemoglobinopathy screening tests such as electrophoresis cannot always diagnose them definitively. The laboratory will take both the blood count and electrophoresis results into account and provide a likely diagnosis.
The most severe form is ß°-thalassaemia, where both ß genes are deleted. This causes anaemia from early childhood – regular transfusion and iron chelation is required. Haemoglobin H disease occurs where three out of four a-chains are deleted, and causes anaemia, splenomegaly, haemolysis and iron overload in childhood or early adulthood.
Click here for a clinical review of the thalassaemias.
The National sickle and thalassaemia screening programme provides antenatal and newborn screening for the most important haemoglobinopathies.
How should haemochromatosis be diagnosed? What screening tests should be used – many of us only have access to ferritin and TIBC. Who should have genetic screening?
About 0.7% of individuals in the UK have genetic haemachromatosis. The increased body iron due to increased absorption causes progressive iron overload causing cardiac, endocrine and liver failure and joint pain. The treatment consists of venesection to remove the excess iron. The diagnosis is made incidentally, during investigation for liver, cardiac or endocrine failure or as part of family studies when a relative is known to be affected or when unexpectedly high levels of iron stores are found during routine investigation.
The most readily available screening test is transferrin saturation. It can be calculated as total iron binding capacity divided by the serum iron. Transferrin saturations of >55% for men or >50% for women require further investigation for haemochromatosis. Transferrin saturation may not be raised in children or young adults, so it is reasonable to delay testing until adulthood.
Ferritin increases when iron overload occurs. Ferritin levels of >300ug/l in men or postmenopausal women and >200ug/l in premenopausal women could be considered raised. But it is also raised in infection, inflammation or hepatitis of any cause. If the ferritin is high, measurement of transferrin saturation is advised.
The current genetic tests for haemochromatosis detect about 95% of causative mutations. Genetic testing is warranted in patients with evidence of iron overload, either biochemically (raised transferrin saturation) or by organ involvement – for example, via liver biopsy or cardiac MRI. It is also reasonable to perform genetic testing on first-degree relatives of patients with haemochromatosis caused by one of the known mutations (C282Y or H63D), even if that relative does not yet have evidence of iron overload. Asymptomatic individuals can be monitored for iron overload with annual transferrin saturation and ferritin measurements.
We would advise against genetic testing for children but they and their parents should be informed of the risk and offered genetic testing when they reach adulthood and can judge the implications for themselves.
In patients with a possible thrombophilia how useful are current screening tests? Who might be at risk that we might not consider? If the history suggests a risk but tests are negative what precautions should be taken?
Thrombophilia testing detects some hereditary and acquired risk factors for venous thromboembolism. Risk factors for atherosclerotic arterial events are different –
hypertension, smoking diabetes and lipids. Antiphospholipid antibodies – lupus anticoagulant and anticardiolipin antibody – should be tested in the occasional patient with an arterial thrombosis at a young age. But this is the only thrombophilic factor that causes arterial events in adults.
The value of thrombophilia testing is much debated by the experts – it may be useful in selected patients to determine the duration and intensity of anticoagulation after a thrombotic event, but clinical factors are also important. Being male and having a spontaneous venous thromboembolism are strong risk factors for recurrent events and predict a high risk of recurrence.
Thrombophilia testing is expensive, the results can be difficult to interpret, and it is of limited clinical value in unselected patients. Common pitfalls include sending thrombophilia screens while the patient is on warfarin – which invalidates most lupus anticoagulant tests – and overinterpreting low protein S levels, which are usually low in high-oestrogen states such as pregnancy and in those on the oral contraceptive pill.
Remember many thrombophilia risk factors (factor V Leiden heterozygosity, prothrombin gene mutation, mild protein S deficiency) are of high prevalence and carry low relative risk of venous thromboembolism, so it is important not to cause undue anxiety.
Thrombophilia screening may be useful in some settings.
Those with a prior DVT should receive thromboprophylaxis during any future hospital admissions or periods of immobility. During any long-distance flights, compression stockings should be worn, leg exercises should be performed regularly (suggested exercises are usually given in the in-flight magazine) and dehydration should be avoided (drink plenty of water and no coffee or alcohol).
Women need advice from an expert about thromboprophylaxis during pregnancy.
Dr Robert Weinkove is specialist registrar and Dr Beverley Hunt is consultant haematologist at Guy's and St Thomas' Trust, London
Dr Colin Kenny is a GP in Dromore, Northern Irelandhaematology NTK referral box thrombophilia screening What I will do now
Dr Kenny reflects on the answers to his questions
• I will consider a diagnosis of polycythaemia when I see a haematocrit of >0.51 (or 51%) for a man or >0.48 (or 48%) for a woman.
• When seeing Afro-Caribbean patients I will bear inWhat I will do now
What I will do now
Dr Kenny reflects on the answers to his questions
• I will consider a diagnosis of polycythaemia when I see a haematocrit of >0.51 (or 51%) for a man or >0.48 (or 48%) for a woman.
• When seeing Afro-Caribbean patients I will bear in mind that they may have lower platelet and neutrophil counts than similar Caucasian patients.
• I will remember sickle cell disease and the thalassaemias may coexist but have important different geographical distributions.
• When initiating hydroxobalamin I will give an initial replacement of six injections of 1mg intramuscular hydroxobalamin over two or three weeks. After this I will not routinely monitor levels.
• When considering myeloma I will send both serum and urine for electrophoresis.
• When I see ferritin levels of >300 ug/l in men or post-menopausal women and >200ug/l in pre-menopausal women I will consider a diagnosis of haemachromatosis.
• I will bear in mind that genetic testing for haemachromatosis has limitations as as well as insurance implications.