The formation of blood cells (haemopoiesis)

The haemopoietic system includes the bone marrow, liver, spleen, lymph nodes and thymus. There is huge turnover of cells with the red cells surviving 120 days, platelets around 7 days but granulocytes only 7 hours. The production of as many as 10¹³ new myeloid cells (all blood cells except for lymphocytes) per day in the normal healthy state requires tight regulation according to the needs of the body.

Blood islands are formed in the yolk sac in the 3rd week of gestation and produce primitive blood cells, which migrate to the liver and spleen. These organs are the chief sites of haemopoiesis from 6 weeks to 7 months, when the bone marrow becomes the main source of blood cells. However, in childhood and adult life, the bone marrow is the only source of blood cells in a normal person.

At birth, haemopoiesis is present in the marrow of nearly every bone. As the child grows, the active red marrow is gradually replaced by fat (yellow marrow) so that haemopoiesis in the adult becomes confined to the central skeleton and the proximal ends of the long bones. Only if the demand for blood cells increases and persists do the areas of red marrow extend. Pathological processes interfering with normal haemopoiesis may result in resumption of haemopoietic activity in the liver and spleen, which is referred to as extramedullary haemopoiesis.

All blood cells are derived from pluripotent stem cells. These stem cells are supported by stromal cells (see below), which also influence haemopoiesis. The stem cell has two properties – the first is self-renewal, i.e. the production of more stem cells, and the second is its proliferation and differentiation into progenitor cells, committed to one specific cell line.

There are two major ancestral cell lines derived from the pluripotential stem cell: lymphocytic and myeloid (non-lymphocytic) cells (Fig. 8.1). The former gives rise to T and B cells. The myeloid stem cell gives rise to the progenitor CFU-GEMM (colony-forming unit, granulocyte–erythrocyte–monocyte–megakaryocyte). The CFU-GEMM can go on to form CFU-GM, CFU-Eo, and CFU-Meg, each of which can produce a particular cell type (i.e. neutrophils, eosinophils and platelets) under appropriate growth conditions. The progenitor cells such as CFU-GEMM cannot be recognized in bone marrow biopsies but are recognized by their ability to form colonies when haemopoietic cells are immobilized in a soft gel matrix. Haemopoiesis is under the control of growth factors and inhibitors, and the microenvironment of the bone marrow also plays a role in its regulation.

Figure 8.1 Role of growth factors in normal haemopoiesis. Some of the multiple growth factors acting on stem cells and early progenitor cells are shown. baso, basophil; BFU, burst-forming unit; CFU, colony-forming unit; CSF, colony-stimulating factor; E, erythroid; Eo, eosinophil; EPO, erythropoietin; G, granulocyte; GEMM, mixed granulocyte, erythroid, monocyte, megakaryocyte; GM, granulocyte, monocyte; IL, interleukin; M, monocyte; Meg, megakaryocyte; SCF, stem cell (Steel) factor or C-kit ligand; TNF, tumour necrosis factor; TPO, thrombopoietin.

Haemopoietic growth factors

Haemopoietic growth factors are glycoproteins, which regulate the differentiation and proliferation of haemopoietic progenitor cells and the function of mature blood cells. They act on the cytokine-receptor superfamily expressed on haemopoietic cells at various stages of development to maintain the haemopoietic progenitor cells and to stimulate increased production of one or more cell lines in response to stresses such as blood loss and infection (Fig. 8.1).

These haemopoietic growth factors including erythropoietin, interleukin 3 (IL-3), IL-6, -7, -11, -12, β-catenin, stem cell factor (SCF, Steel factor or C-kit ligand) and Fms-tyrosine kinase 3 (Flt3) act via their specific receptor on cell surfaces to stimulate the cytoplasmic janus kinase (JAK) (see p. 25). This major signal transducer activates tyrosine kinase causing gene activation in the cell nucleus. Colony-stimulating factors (CSFs, the prefix indicating the cell type, see Fig. 8.1), as well as interleukins and erythropoietin (EPO) regulate the lineage committed progenitor cells.

Thrombopoietin (TPO, which, like erythropoietin, is produced in the kidneys and the liver) controls platelet production, along with IL-6 and IL-11. In addition to these factors stimulating haemopoiesis, other factors inhibit the process and include tumour necrosis factor (TNF) and transforming growth factor-β (TGF-β). Many of the growth factors are produced by activated T cells, monocytes and bone marrow stromal cells such as fibroblasts, endothelial cells and macrophages; these cells are also involved in inflammatory responses. Bone marrow stem cells can differentiate into other organ cell types, e.g. heart, liver, nerves, bone and this is called stem cell plasticity.

Peripheral blood

Automated cell counters are used to measure the haemoglobin concentration (Hb) and the number and size of red cells, white cells and platelets (Table 8.1). Other indices can be derived from these values. A carefully evaluated blood film is still an essential adjunct to the above, as definitive abnormalities of cells can be seen.

Table 8.1 Normal values for peripheral blood

ESR, erythrocyte sedimentation rate; Hb, haemoglobin; MCH, mean corpuscular haemoglobin; MCHC, mean corpuscular haemoglobin concentration; MCV, mean corpuscular volume of red cells; PCV, packed cell volume; RCC, red cell count; RDW, red blood cell distribution width; WBC, white blood count.

Erythropoiesis

Red cell precursors pass through several stages in the bone marrow. The earliest morphologically recognizable cells are pronormoblasts. Smaller normoblasts result from cell divisions, and precursors at each stage progressively contain less RNA and more Hb in the cytoplasm. The nucleus becomes more condensed and is eventually lost from the late normoblast in the bone marrow when the cell becomes a reticulocyte.

Haemoglobin synthesis

Haemoglobin performs the main functions of red cells – carrying O₂ to the tissues and returning CO₂ from the tissues to the lungs. Each normal adult Hb molecule (HbA) has a molecular weight of 68 000 and consists of two α and two β globin polypeptide chains (α₂β₂). HbA comprises about 97% of the Hb in adults. Two other haemoglobin types, HbA₂ (α₂δ₂) and HbF (α₂γ₂), are found in adults in small amounts (1.5–3.2% and <1%, respectively) (see p. 390).

Haemoglobin synthesis occurs in the mitochondria of the developing red cell (Fig. 8.2). The major rate-limiting step is the conversion of glycine and succinic acid to δ-aminolaevulinic acid (ALA) by ALA synthase. Vitamin B₆ is a coenzyme for this reaction, which is inhibited by haem and stimulated by erythropoietin. Two molecules of δ-ALA condense to form a pyrrole ring (porphobilinogen). These rings are then grouped in fours to produce protoporphyrins and with the addition of iron haem is formed. Haem is then inserted into the globin chains to form a haemoglobin molecule. The structure of Hb is shown in Figure 8.3.

Figure 8.2 Haemoglobin synthesis. Transferrin attaches to a surface receptor on developing red cells. Iron is released and transported to the mitochondria, where it combines with protoporphyrin to form haem. Protoporphyrin itself is manufactured from glycine and succinyl-CoA. Haem combines with α and β chains (formed on ribosomes) to make haemoglobin.

Figure 8.3 Model of the haemoglobin molecule showing α (red) and β (blue) chains. 2,3-BPG (bisphosphoglycerate) binds in the centre of the molecule and stabilizes the deoxygenated form by cross-linking the β chains (also see Fig. 8.4). M, methyl; P, propionic acid; V, vinyl.

Haemoglobin function

The biconcave shape of red cells provides a large surface area for the uptake and release of oxygen and carbon dioxide. Haemoglobin becomes saturated with oxygen in the pulmonary capillaries where the partial pressure of oxygen is high and Hb has a high affinity for oxygen. Oxygen is released in the tissues where the partial pressure of oxygen is low and Hb has a low affinity for oxygen.

In adult haemoglobin (Hb), a haem group is bound to each of the four globin chains; the haem group has a porphyrin ring with a ferrous atom which can reversibly bind one oxygen molecule. The haemoglobin molecule exists in two conformations, R and T. The T (taut) conformation of deoxyhaemoglobin is characterized by the globin units being held tightly together by electrostatic bonds (Fig. 8.4). These bonds are broken when oxygen binds to haemoglobin, resulting in the R (relaxed) conformation in which the remaining oxygen binding sites are more exposed and have a much higher affinity for oxygen than in the T conformation. The binding of one oxygen molecule to deoxyhaemoglobin increases the oxygen affinity of the remaining binding sites – this property is known as ‘cooperativity’ and is the reason for the sigmoid shape of the oxygen dissociation curve. Haemoglobin is, therefore, an example of an allosteric protein. The binding of oxygen can be influenced by secondary effectors – hydrogen ions, carbon dioxide and red-cell 2,3-bisphosphoglycerate (2,3-BPG). Hydrogen ions and carbon dioxide added to blood cause a reduction in the oxygen-binding affinity of haemoglobin (the Bohr effect). Oxygenation of haemoglobin reduces its affinity for carbon dioxide (the Haldane effect). These effects help the exchange of carbon dioxide and oxygen in the tissues.

Figure 8.4 Oxygenated and deoxygenated haemoglobin molecule. The haemoglobin molecule is predominantly stabilized by α-β chain bonds rather than α-α and β-β chain bonds. The structure of the molecule changes during O₂ uptake and release. When O₂ is released, the β chains rotate on the α₁β₂ and α₂β₁ contacts, allowing the entry of 2,3-BPG which causes a lower affinity of haemoglobin for O₂ and improved delivery of O₂ to the tissues.

Red cell metabolism produces 2,3-BPG from glycolysis. 2,3-BPG accumulates because it is sequestered by binding to deoxyhaemoglobin. The binding of 2,3-BPG stabilizes the T conformation and reduces its affinity for oxygen. The P₅₀ is the partial pressure of oxygen at which the haemoglobin is 50% saturated with oxygen. P₅₀ increases with 2,3-BPG concentrations, which increase when oxygen availability is reduced in conditions such as hypoxia or anaemia. P₅₀ also rises with increasing body temperature, which may be beneficial during prolonged exercise. Haemoglobin regulates oxygen transport as shown in the oxyhaemoglobin dissociation curve. When the primary limitation to oxygen transport is in the periphery, e.g. heavy exercise, anaemia, the P₅₀ is increased to enhance oxygen unloading. When the primary limitation is in the lungs, e.g. lung disease, high altitude exposure, the P₅₀ is reduced to enhance oxygen loading.

A summary of normal red cell production and destruction is given in Figure 8.5.

Figure 8.5 Red cell production and breakdown (see p. 307).

Iron

Absorption

Factors influencing iron and haem iron absorption (Fig. 8.8) are shown in Table 8.2.

Figure 8.8 (a) Regulation of the absorption of intestinal iron. The iron-absorbing cells of the duodenal epithelium originate in the intestinal crypts and migrate toward the tip of the villus as they differentiate (maturation axis). Absorption of intestinal iron is regulated by at least three independent mechanisms, although the protein hepcidin is key. First, iron absorption is influenced by recent dietary iron intake (dietary regulator). After a large dietary bolus, absorptive cells are resistant to iron uptake for several days. Second, iron absorption can be modulated considerably in response to body iron stores (stores regulator). Third, a signal communicates the state of bone marrow erythropoiesis to the intestine (erythroid regulator). (b) Duodenal crypt cells sense body iron status through the binding of transferrin to the HFE/B₂M/TfR1 gene complex. Cytosolic enzymes change the oxidative state of iron from ferric (Fe³⁺) to ferrous (Fe²+). A decrease in crypt cell iron concentration upregulates the divalent metal transporter (DMT1). This increases as crypt cells migrate up the villus and become mature absorptive cells. (c) Apical cell. Dietary iron is reduced from the ferric to the ferrous state by the brush border ferrireductase. DMT1 facilitates iron absorption from the intestinal lumen. The export proteins, e.g. ferroportin 1 and hephaestin, transfer iron from the enterocyte into the circulation depending on the hepcidin level. A second pathway absorbs intact haem iron into the circulation via BRCP and FLVCR. BCRP, breast cancer resistant protein; B2M, β₂-microglobulin; FLVCR, feline leukaemia virus subgroup C; HCP1, haem carrier protein-1; HFE, hereditary haemochromatosis gene; TfR1, transferrin receptor.

Table 8.2 Factors influencing iron absorption

Haem iron is absorbed better than non-haem iron Ferrous iron is absorbed better than ferric iron Gastric acidity helps to keep iron in the ferrous state and soluble in the upper gut Formation of insoluble complexes with phytate or phosphate decreases iron absorption Iron absorption is increased with low iron stores and increased erythropoietic activity, e.g. bleeding, haemolysis, high altitude There is a decreased absorption in iron overload, except in hereditary haemochromatosis, where it is increased

Dietary haem iron is more rapidly absorbed than non-haem iron derived from vegetables and grain. Most haem is absorbed in the proximal intestine, with absorptive capacity decreasing distally. The intestinal haem transporter HCP1 (haem carrier protein 1) has been identified and found to be highly expressed in the duodenum. It is upregulated by hypoxia and iron deficiency. Some haem iron may be reabsorbed intact into circulation via the cell by two exporter proteins – BCRP (breast cancer resistant protein) and FLVCR (feline leukaemia virus subgroup C) (Fig. 8.8).

Non-haem iron absorption occurs primarily in the duodenum. Non-haem iron is dissolved in the low pH of the stomach and reduced from the ferric to the ferrous form by a brush border ferrireductase. Cells in duodenal crypts are able to sense the body’s iron requirements and retain this information as they mature into cells capable of absorbing iron at the tips of the villi. A protein, divalent metal transporter 1 (DMT1) or natural resistance-associated macrophage protein (NRAMP2), transports iron (and other metals) across the apical (luminal) surface of the mucosal cells in the small intestine.

Once inside the mucosal cell, iron may be transferred across the cell to reach the plasma, or be stored as ferritin; the body’s iron status at the time the absorptive cell developed from the crypt cell is probably the crucial deciding factor. Iron stored as ferritin will be lost into the gut lumen when the mucosal cells are shed; this regulates iron balance. The mechanism of transport of iron across the basolateral surface of mucosal cells involves a transporter protein, ferroportin 1 (FPN 1) through its iron-responsive element (IRE). This transporter protein requires an accessory, multicopper protein, hephaestin (Fig. 8.8).

The body iron content is closely regulated by the control of iron absorption but there is no physiological mechanism for eliminating excess iron from the body. The key molecule regulating iron absorption is hepcidin, a 25 amino acid peptide synthesized in the liver. Hepcidin acts by regulating the activity of the iron exporting protein ferroportin by binding to ferroportin causing its internalization and degradation, thereby decreasing iron efflux from iron exporting tissues into plasma. Therefore, high levels of hepcidin (occurring in inflammation states) via inflammatory cytokines, e.g. IL-6 will destroy ferroportin and limit iron absorption, and low levels of hepcidin (e.g. in anaemia, low iron stores, hypoxia) will encourage iron absorption. For example, in patients with haemochromatosis, mutations in the genes HFE, HJV and TfR2 will interrupt hepcidin synthesis. Therefore, in the intestinal cells, a deficiency of hepcidin would lead to less ferroportin being bound and thus more iron will be released into the plasma.

A longstanding mystery is why anaemias characterized by ineffective erythropoiesis such as thalassaemia are associated with excessive and inappropriate iron absorption. Preliminary evidence again suggests that the increased iron absorption in β-thalassaemia is mediated by downregulation of hepcidin and upregulation of ferroportin.

FURTHER READING

Andrews NC. Closing the iron gate. N Engl J Med 2012; 366:376−377.

Fleming RE, Ponka P. Iron overload in human disease. N Engl J Med 2012; 366:1548−1550.

Hentze MW, Muckenthaler MU, Galy B et al. Two to tango: regulation of mammalian iron absorption. Cell 2010; 142:24–38.

Iron deficiency

Iron deficiency anaemia develops when there is inadequate iron for haemoglobin synthesis. The causes are:

Most iron deficiency is due to blood loss, usually from the uterus or gastrointestinal tract. Premenopausal women are in a state of precarious iron balance owing to menstruation. A common cause of iron deficiency worldwide is blood loss from the gastrointestinal tract resulting from parasites such as hookworm infestation. The poor quality of the diet, predominantly containing vegetables, also contributes to the high prevalence of iron deficiency in developing countries. Even in developed countries, iron deficiency is not uncommon in infancy where iron intake is insufficient for the demands of growth. It is more prevalent in infants born prematurely or where the introduction of mixed feeding is delayed.

Clinical features

The symptoms of anaemia are described on page 375. The well known clinical features of iron deficiency listed below are generally only seen in cases of very longstanding iron deficiency:

Brittle nails

Spoon-shaped nails (koilonychia)

Atrophy of the papillae of the tongue

Angular stomatitis

Brittle hair

A syndrome of dysphagia and glossitis (Plummer–Vinson or Paterson–Brown–Kelly syndrome; see p. 243).

The diagnosis of iron deficiency anaemia relies on a clinical history which should include questions about dietary intake, self-medication with non-steroidal anti-inflammatory drugs (which may give rise to gastrointestinal bleeding), and the presence of blood in the faeces (which may be a sign of haemorrhoids or carcinoma of the lower bowel). In women, a careful enquiry about the duration of periods, the occurrence of clots and the number of sanitary towels or tampons (normal 3–5/day) used should be made.

Investigations

Blood count and film. A characteristic blood film is shown in Figure 8.9. The red cells are microcytic (MCV <80 fL) and hypochromic (MCH (mean corpuscular haemoglobin) <27 pg). There is poikilocytosis (variation in shape) and anisocytosis (variation in size). Target cells are seen.

Serum iron and iron-binding capacity. The serum iron falls and the total iron-binding capacity (TIBC) rises in iron deficiency compared with normal. Iron deficiency is regularly present when the transferrin saturation (i.e. serum iron divided by TIBC) falls below 19% (Table 8.3).

Serum ferritin. The level of serum ferritin reflects the amount of stored iron. The normal values for serum ferritin are 30–300 µg/L (11.6–144 nmol/L) in males and 15–200 µg/L (5.8–96 nmol/L) in females. In simple iron deficiency, a low serum ferritin confirms the diagnosis. However, ferritin is an acute-phase reactant, and levels increase in the presence of inflammatory or malignant diseases. Very high levels of ferritin may be observed in hepatitis and in a rare disease, haemophagocytic lymphohistiocytosis (p. 80).

Serum soluble transferrin receptors. The number of transferrin receptors increases in iron deficiency. The results of this immunoassay compare well with results from bone marrow aspiration at estimating iron stores. This assay can help to distinguish between iron deficiency and anaemia of chronic disease (Table 8.3), and may avoid the need for bone marrow examination even in complex cases. It may sometimes be helpful in the investigation of complicated causes of anaemia.

Other investigations. These will be indicated by the clinical history and examination. Investigations of the gastrointestinal tract are often required to determine the cause of the iron deficiency (see p. 257).

Figure 8.9 Hypochromic microcytic cells (arrow) on a blood film. Poikilocytosis (variation in size) and anisocytosis (variation in shape) are seen.

Table 8.3 Microcytic anaemia: the differential diagnosis

Differential diagnosis

The presence of anaemia with microcytosis and hypochromia does not necessarily indicate iron deficiency. The most common other causes are thalassaemia, sideroblastic anaemia and anaemia of chronic disease, and in these disorders the iron stores are normal or increased. The differential diagnosis of microcytic anaemia is shown in Table 8.3.

Treatment

The correct management of iron deficiency is to find and treat the underlying cause, and to give iron to correct the anaemia and replace iron stores. Patients with iron deficiency taking iron will increase their Hb level by approximately 10 g/L/week unless of course other factors such as bleeding are present.

Oral iron is all that is required in most cases. The best preparation is ferrous sulphate (200 mg three times daily, a total of 180 mg ferrous iron), which is absorbed best when the patient is fasting. If the patient has side-effects such as nausea, diarrhoea or constipation, taking the tablets with food or reducing the dose using a preparation with less iron such as ferrous gluconate (300 mg twice daily, only 70 mg ferrous iron) is all that is usually required to reduce the symptoms.

In developing countries, distribution of iron tablets and fortification of food are the main approaches for the alleviation of iron deficiency. However, iron supplementation programmes have been ineffective, probably mainly because of poor compliance.

Oral iron should be given for long enough to correct the Hb level and to replenish the iron stores; this can take 6 months. The commonest causes of failure to respond to oral iron are:

Lack of compliance

Continuing haemorrhage

Incorrect diagnosis, e.g. thalassaemia trait.

These possibilities should be considered before parenteral iron is used. However, parenteral iron is required by occasional patients, e.g. intolerant to oral preparation, severe malabsorption, chronic disease (e.g. inflammatory bowel disease). Iron stores are replaced much faster with parenteral iron than with oral iron, but the haematological response is no quicker. Parenteral iron can be given by slow intravenous infusion of low-molecular-weight iron dextran (test dose required), iron sucrose, ferric carboxymaltose, iron isomaltoside 1000; oral iron should be discontinued.

Anaemia of chronic disease

One of the most common types of anaemia, particularly in hospital patients, is the anaemia of chronic disease, occurring in patients with chronic infections such as tuberculosis or chronic inflammatory disease such as Crohn’s disease, rheumatoid arthritis, systemic lupus erythematosus (SLE), polymyalgia rheumatica and malignant disease. There is decreased release of iron from the bone marrow to developing erythroblasts, an inadequate erythropoietin response to the anaemia, and decreased red cell survival.

The exact mechanisms responsible for these effects are not clear but it seems likely that high levels of hepcidin expression play a key role (see iron absorption).

The serum iron and the TIBC are low, and the serum ferritin is normal or raised because of the inflammatory process. The serum soluble transferrin receptor level is normal (Table 8.3). Stainable iron is present in the bone marrow, but iron is not seen in the developing erythroblasts. Patients do not respond to iron therapy, and treatment is, in general, that of the underlying disorder. Recombinant erythropoietin therapy is used in the anaemia of renal disease (see p. 623), and occasionally in inflammatory disease (rheumatoid arthritis, inflammatory bowel disease).

Sideroblastic anaemia

Sideroblastic anaemias are inherited or acquired disorders characterized by a refractory anaemia, a variable number of hypochromic cells in the peripheral blood, and excess iron and ring sideroblasts in the bone marrow. The presence of ring sideroblasts is the diagnostic feature of sideroblastic anaemia. There is accumulation of iron in the mitochondria of erythroblasts owing to disordered haem synthesis forming a ring of iron granules around the nucleus that can be seen with Perls’ reaction. The blood film is often dimorphic; ineffective haem synthesis is responsible for the microcytic hypochromic cells. Sideroblastic anaemias can be inherited as an X-linked disease transmitted by females. Acquired causes include myelodysplasia, myeloproliferative disorders, myeloid leukaemia, drugs (e.g. isoniazid), alcohol misuse and lead toxicity. It can also occur in other disorders such as rheumatoid arthritis, carcinomas, megaloblastic and haemolytic anaemias. A structural defect in δ-aminolaevulinic acid (ALA) synthase, the pyridoxine-dependent enzyme responsible for the first step in haem synthesis (Fig. 8.2), has been identified in one form of inherited sideroblastic anaemia. Primary acquired sideroblastic anaemia is one of the myelodysplastic syndromes (see p. 405) and this is the cause of the vast majority of cases of sideroblastic anaemia in adults. Lead toxicity is described in Chapter 17.

Vitamin B₁₂

Vitamin B₁₂ is synthesized by certain microorganisms, and humans are ultimately dependent on animal sources. It is found in meat, fish, eggs and milk, but not in plants. Vitamin B₁₂ is not usually destroyed by cooking. The average daily diet contains 5–30 µg of vitamin B₁₂, of which 2–3 µg is absorbed. The average adult stores some 2–3 mg, mainly in the liver, and it may take 2 years or more after absorptive failure before B₁₂ deficiency develops, as the daily losses are small (1–2 µg).

Structure and function

Cobalamins consist of a planar group with a central cobalt atom (corrin ring) and a nucleotide set at right-angles (Fig. 8.13). Vitamin B₁₂ was first crystallized as cyanocobalamin, but the main natural cobalamins have deoxyadenosyl-, methyl- and hydroxocobalamin groups attached to the cobalt atom.

Figure 8.13 Methylcobalamin structure. This is the main form of vitamin B₁₂ in the plasma.

The main function of B₁₂ is the methylation of homocysteine to methionine with the demethylation of methyl THF polyglutamate to THF. THF is a substrate for folate polyglutamate synthesis.

Deoxyadenosylcobalamin is a coenzyme for the conversion of methylmalonyl CoA to succinyl CoA. Measurement of methylmalonic acid in urine was used as a test for vitamin B₁₂ deficiency but it is no longer carried out routinely.

Pernicious anaemia

Pernicious anaemia (PA) is an autoimmune disorder in which there is atrophic gastritis with loss of parietal cells in the gastric mucosa with consequent failure of intrinsic factor production and vitamin B₁₂ malabsorption.

Folic acid

Folic acid monoglutamate is not itself present in nature but occurs as polyglutamates. Folates are present in food as polyglutamates in the reduced dihydrofolate or tetrahydrofolate (THF) forms (Fig. 8.14), with methyl (CH₃), formyl (CHO) or methylene (CH₂) groups attached to the pteridine part of the molecule. Polyglutamates are broken down to monoglutamates in the upper gastrointestinal tract, and during the absorptive process these are converted to methyl THF monoglutamate, which is the main form in the serum. The methylation of homocysteine to methionine requires both methylcobalamin and methyl THF as coenzymes. This reaction is the first step in which methyl THF entering cells from the plasma is converted into folate polyglutamates. Intracellular polyglutamates are the active forms of folate and act as coenzymes in the transfer of single carbon units in amino acid metabolism and DNA synthesis (Fig. 8.12).

Figure 8.14 Folic acid structure. This is formed from three building blocks as shown. Tetrahydrofolate has additional hydrogen atoms at positions 5, 6, 7 and 8.

Treatment and prevention of megaloblastic anaemia

Treatment depends on the type of deficiency. Blood transfusion is not indicated in chronic anaemia; indeed, it is dangerous to transfuse elderly patients, as heart failure may be precipitated. Folic acid may produce a haematological response in vitamin B₁₂ deficiency but may aggravate the neuropathy. Large doses of folic acid alone should not be used to treat megaloblastic anaemia unless the serum vitamin B₁₂ level is known to be normal. In severely ill patients, it may be necessary to treat with both folic acid and vitamin B₁₂ while awaiting serum levels.

Treatment of vitamin B₁₂ deficiency

Hydroxocobalamin 1000 µg can be given intramuscularly to a total of 5–6 mg over the course of 3 weeks; 1000 µg is then necessary every 3 months for the rest of the patient’s life. Alternatively, oral B₁₂ 2 mg/day is used, as 1–2% of an oral dose is absorbed by diffusion and therefore does not require intrinsic factor.

Compliance with an oral daily regimen may be a problem, particularly in elderly patients. The use of sublingual nuggets of B₁₂ (2 × 1000 µg daily) has been suggested to be an effective and more convenient option.

Clinical improvement may occur within 48 hours and a reticulocytosis can be seen some 2–3 days after starting therapy, peaking at 5–7 days. Improvement of the polyneuropathy may occur over 6–12 months, but longstanding spinal cord damage is irreversible. Hypokalaemia can occur and, if severe, supplements should be given. Iron deficiency often develops in the first few weeks of therapy. Hyperuricaemia also occurs but clinical gout is uncommon. In patients who have had a total gastrectomy or an ileal resection, vitamin B₁₂ should be monitored; if low levels occur, prophylactic vitamin B₁₂ should be given. Vegans may require B₁₂ supplements.

Evidence for haemolysis

Increased red cell breakdown is accompanied by increased red cell production. This is shown in Figure 8.16.

Hereditary spherocytosis (HS)

HS is the most common inherited haemolytic anaemia in northern Europeans, affecting 1 in 5000. It is inherited in an autosomal dominant manner, but in 25% of patients, neither parent is affected and it is presumed that HS has occurred by spontaneous mutation or is truly recessive. HS is due to defects in the red cell membrane, resulting in the cells losing part of the cell membrane as they pass through the spleen, possibly because the lipid bilayer is inadequately supported by the membrane skeleton. The best-characterized defect is a deficiency in the structural protein spectrin, but quantitative defects in other membrane proteins have been identified (Fig. 8.18), with ankyrin defects being the most common. The abnormal red cell membrane in HS is associated functionally with an increased permeability to sodium, and this requires an increased rate of active transport of sodium out of the cells which is dependent on ATP produced by glycolysis. The surface-to-volume ratio decreases, and the cells become spherocytic. Spherocytes are more rigid and less deformable than normal red cells. They are unable to pass through the splenic microcirculation so they have a shortened lifespan.

Investigations

Anaemia. This is usually mild, but occasionally can be severe.

Blood film. This shows spherocytes (Fig. 8.19) and reticulocytes.

Haemolysis is evident (e.g. the serum bilirubin and urinary urobilinogen will be raised).

Osmotic fragility. When red cells are placed in solutions of increasing hypotonicity, they take in water, swell, and eventually lyse. Spherocytes tolerate hypotonic solutions less well than do normal biconcave red cells. Osmotic fragility tests are infrequently carried out in routine practice, but may be useful to confirm a suspicion of spherocytosis on a blood film.

Direct antiglobulin (Coombs’) test is negative in hereditary spherocytosis, virtually ruling out autoimmune haemolytic anaemia where spherocytes are also commonly present.

Figure 8.19 Spherocytes (arrowed). This blood film also shows reticulocytes, polychromasia and a nucleated erythroblast.

Hereditary elliptocytosis

This disorder of the red cell membrane is inherited in an autosomal dominant manner and has a prevalence of 1 in 2500 in Caucasians. The red cells are elliptical due to deficiencies of protein 4.1 or the spectrin/actin/4.1 complex which leads to weakness of the horizontal protein interaction and to the membrane defect (Fig. 8.18). Clinically it is a similar condition to HS but milder. Only a minority of patients have anaemia and only occasional patients require splenectomy.

Rarely, hereditary spherocytosis or elliptocytosis may be inherited in a homozygous fashion giving rise to a severe haemolytic anaemia sometimes necessitating splenectomy in early childhood.

Hereditary stomatocytosis

Stomatocytes are red cells in which the pale central area appears slit-like. Their presence in large numbers may occur in a hereditary haemolytic anaemia associated with a membrane defect, but excess alcohol intake is also a common cause. Although these hereditary conditions are very rare, a correct diagnosis is required since splenectomy is contraindicated as it may result in fatal thromboembolic events.

β-Thalassaemia

In homozygous β-thalassaemia, either no normal β chains are produced (β⁰) or β-chain production is very reduced (β⁺). There is an excess of α chains, which precipitate in erythroblasts and red cells causing ineffective erythropoiesis and haemolysis. The excess α chains combine with whatever β, δ and γ chains are produced, resulting in increased quantities of HbA₂ and HbF and, at best, small amounts of HbA. In heterozygous β-thalassaemia there is usually symptomless microcytosis with or without mild anaemia. Table 8.10 shows the findings in the homozygote and heterozygote for the common types of β-thalassaemia.

Table 8.10 β-Thalassaemia: common findings

Hb Lepore is a cross-fusion product of δ and β globin genes.

Adapted with permission from Weatherall DJ. Disorders of the synthesis of function of haemoglobin. In: Weatherall DJ, Warrell DA, Cox TM, Firth JD (eds) Oxford Textbook of Medicine, 5th edn. Oxford: Oxford University Press; 2010.

Molecular genetics

The molecular errors accounting for over 200 genetic defects leading to β-thalassaemia have been characterized. Unlike in α-thalassaemia, the defects are mainly point mutations rather than gene deletions. The mutations result in defects in transcription, RNA splicing and modification, translation via frame shifts and nonsense codons producing highly unstable β-globin, which cannot be utilized.

Clinical syndromes

Clinically, β-thalassaemia can be divided into the following:

Thalassaemia minor (or trait), the symptomless heterozygous carrier state

Thalassaemia intermedia, a moderate anaemia, not requiring regular transfusions (with a number of different genotypes)

Thalassaemia major (generally homozygous β-thalassaemia), severe anaemia requiring regular transfusions.

Thalassaemia minor (trait)

This common carrier state (heterozygous β-thalassaemia) is asymptomatic. Anaemia is mild or absent. The red cells are hypochromic and microcytic with a low MCV and MCH, and it may be confused with iron deficiency. However, the two are easily distinguished, as in thalassaemia trait the serum ferritin and the iron stores are normal (Table 8.2). The RDW is usually normal (see p. 373). Hb electrophoresis usually shows a raised HbA₂ and often a raised HbF (Fig. 8.22). Iron should not be given to these patients unless they also have proven coincidental iron deficiency.

Figure 8.22 Patterns of haemoglobin electrophoresis.

Thalassaemia intermedia

Thalassaemia intermedia includes patients who are symptomatic with moderate anaemia (Hb 70–100 g/L), who do not require regular transfusions.

Thalassaemia intermedia may be due to a combination of homozygous mild β+- and α-thalassaemia, where there is reduced α-chain precipitation and less ineffective erythropoiesis and haemolysis. The inheritance of hereditary persistence of HbF with homozygous β-thalassaemia also results in a milder clinical picture than unmodified β-thalassaemia major because the excess α chains are partially removed by the increased production of γ chains.

Patients may have splenomegaly and bone deformities. Recurrent leg ulcers, gallstones and infections are also seen. It should be noted that these patients may be iron overloaded despite a lack of regular blood transfusions. This is caused by excessive iron absorption which results from the underlying dyserythropoiesis (see iron absorption, p. 377).

Thalassaemia major (Cooley’s anaemia)

Most children affected by homozygous β-thalassaemia present during the first year of life with:

Failure to thrive and recurrent bacterial infections

Severe anaemia from 3 to 6 months when the switch from γ- to β-chain production should normally occur

Extramedullary haemopoiesis that soon leads to hepatosplenomegaly and bone expansion, giving rise to the classical thalassaemic facies (Fig. 8.23a).

Figure 8.23 Thalassaemia. (a) A child with thalassaemia, showing the typical facial features. (b) Skull X-ray of a child with β-thalassaemia, showing the ‘hair on end’ appearance. (c) X-ray of hand, showing expansion of the marrow and a thinned cortex.

Skull X-rays in these children show the characteristic ‘hair on end’ appearance of bony trabeculation as a result of expansion of the bone marrow into cortical bone (Fig. 8.23b). The expansion of the bone marrow is also shown in an X-ray of the hand (Fig. 8.23c).

The classic features of untreated thalassaemia major are generally only observed in patients from countries without good blood transfusion support.

FURTHER READING

Higgs DR, Engel JD, Stamatoyannopoulos G. Thalassaemia. Lancet 2012;379:373–383.

Management

The aims of treatment are to suppress ineffective erythropoiesis, prevent bony deformities and allow normal activity and development.

Long-term folic acid supplements are required.

Regular transfusions should be given to keep the Hb above 100 g/L. Blood transfusions may be required every 4–6 weeks.

If transfusion requirements increase, splenectomy may help, although this is usually delayed until after the age of 6 years because of the risk of infection. Prophylaxis against infection is required for patients undergoing splenectomy (see p. 406).

Iron overload caused by repeated transfusions (transfusion haemosiderosis) may lead to damage to the endocrine glands, liver, pancreas and the myocardium by the time patients reach adolescence. Magnetic resonance imaging (myocardial T₂– relaxation time) is useful for monitoring iron overload in thalassaemia; both the heart and the liver can be monitored. The standard iron-chelating agent remains desferrioxamine, although it has to be administered parenterally. Desferrioxamine is given as an overnight subcutaneous infusion on 5–7 nights each week. Ascorbic acid 200 mg daily is given, as it increases the urinary excretion of iron in response to desferrioxamine. Often young children have a very high standard of chelation as it is organized by their parents. However, when the children become adults and take on this role themselves they often rebel and chelation with desferrioxamine may become problematic. Deferiprone, an oral iron chelator, has been available for some years, and results on a new once-daily oral iron chelator, deferasirox, indicate that it is safe, similar in effectiveness to desferrioxamine and is being increasingly used.

Intensive treatment with desferrioxamine has been reported to reverse damage to the heart in patients with severe iron overload, but excessive doses of desferrioxamine may cause cataracts, retinal damage and nerve deafness. Infection with Yersinia enterocolitica occurs in iron-loaded patients treated with desferrioxamine. Iron overload should be periodically assessed by measuring the serum ferritin and by assessment of hepatic iron stores by MRI.

Bone marrow transplantation has been used in young patients with HLA-matched siblings. It has been successful in patients in good clinical condition with a 3-year mortality of <5%, but there is a high mortality (>50%) in patients in poor condition with iron overload and liver dysfunction.

Prenatal diagnosis and gene therapy are discussed on page 43.

Patients’ partners should be tested. If both partners have β-thalassaemia trait, there is a one in four chance of such pregnancy resulting in a child having β-thalassaemia major. Therefore, couples in this situation must be offered prenatal diagnosis (see p. 43).

α-Thalassaemia

Sickle cell anaemia

Sickle cell trait

These individuals have no symptoms unless extreme circumstances cause anoxia, such as flying in non-pressurized aircraft. Sickle cell trait gives some protection against Plasmodium falciparum malaria (see p. 144), and consequently the sickle gene has been seen as an example of a balanced polymorphism (where the advantage of the malaria protection in the heterozygote is balanced by the mortality of the homozygous condition). Typically there is 60% HbA and 40% HbS. It should be emphasized that unlike thalassaemia trait, the blood count and film of a person with sickle cell trait are normal. The diagnosis is made by a positive sickle test or by Hb electrophoresis (Fig. 8.22).

Other structural globin chain defects

There are very many Hb variants and most are not associated with any clinical manifestations. However, some Hb variants may interact with HbS, e.g. compound heterozygosity for HbC and HbS gives rise to HbSC disease. The clinical course of HbSC disease is generally somewhat milder than that of HbSS disease, but there is an increased likelihood of thrombosis, which may lead to thrombosis in pregnancy and to retinopathy.

Combined defects of globin chain production and structure

Abnormalities of Hb structure (e.g. HbS, C) can occur in combination with thalassaemia. The combination of β-thalassaemia trait and sickle cell trait (sickle cell β-thalassaemia) resembles sickle cell anaemia (HbSS) clinically.

HbE (α₂β₂^{26glu → lys}) is the most common Hb variant in South-east Asia, and the second most prevalent haemoglobin variant worldwide. HbE heterozygotes are asymptomatic; the haemoglobin level is normal, but red cells are microcytic. Homozygous HbE causes a mild microcytic anaemia, but the combination of heterozygosity for HbE and β-thalassaemia produces a variable anaemia, which can be as severe as β-thalassaemia major.

Prenatal screening and diagnosis of severe haemoglobin abnormalities

Of the offspring of parents who both have either β-thalassaemia or sickle cell trait, 25% will have β-thalassaemia major or sickle cell anaemia, respectively. Recognition of these heterozygous states in parents and family counselling provide a basis for antenatal screening and diagnosis (p. 44).

Glucose-6-phosphate dehydrogenase (G6PD) deficiency

The enzyme G6PD holds a vital position in the hexose monophosphate shunt (Fig. 8.25), oxidizing glucose-6-phosphate to 6-phosphoglycerate with the reduction of NADP to NADPH. The reaction is necessary in red cells where it is the only source of NADPH, which is used via glutathione to protect the red cell from oxidative damage. G6PD deficiency is a common condition that presents with a haemolytic anaemia and affects millions of people throughout the world, particularly in Africa, around the Mediterranean, the Middle East (around 20%) and South-east Asia (up to 40% in some regions).

The gene for G6PD is localized to chromosome Xq28 near the factor VIII gene. The deficiency is more common in males than in females. However, female heterozygotes can also have clinical problems due to lyonization, whereby because of random X-chromosome inactivation female heterozygotes have two populations of red cells – a normal one and a G6PD-deficient one.

There are over 400 structural types of G6PD, and mutations are mostly single amino acid substitutions (missense point mutations). WHO has classified variants by the degree of enzyme deficiency and severity of haemolysis. The most common types with normal activity are called type B⁺, which is present in almost all Caucasians and about 70% of black Africans, and type A⁺, which is present in about 20% of black Africans. There are many variants with reduced activity but only two are common. In the African, or A⁻ type, the degree of deficiency is mild and more marked in older cells. Haemolysis is self-limiting as the young red cells newly produced by the bone marrow have nearly normal enzyme activity. However, in the Mediterranean type, both young and old red cells have very low enzyme activity. After an oxidant shock the Hb level may fall precipitously; death may follow unless the condition is recognized and the patient is transfused urgently.

Clinical syndromes

Acute drug-induced haemolysis (Table 8.13) – usually dose-related

Favism (ingestion of fava beans)

Chronic haemolytic anaemia

Neonatal jaundice

Infections and acute illnesses will also precipitate haemolysis in patients with G6PD deficiency.

Table 8.13 Drugs causing haemolysis in glucose-6-phosphate deficiency

Analgesics, such as: Aspirin Phenacetin (withdrawn in the UK) Antimalarials, such as: Primaquine Pyrimethamine Quinine Chloroquine Pamaquin Antibacterials, such as: Most sulphonamides Nitrofurantoin Chloramphenicol Quinolones Miscellaneous drugs, such as: Dapsone Vitamin K Probenecid Quinidine Dimercaprol Phenylhydrazine Methylene blue

Mothballs containing naphthalene can also cause haemolysis.

The clinical features are due to rapid intravascular haemolysis with symptoms of anaemia, jaundice and haemoglobinuria.

FURTHER READING

Abboud MR. Hematopoietic stem-cell transplantation for adults with sickle cell disease. N Engl J Med 2009; 361:2380–2381.

Cappellini MD, Fiorelli G. Glucose-6-phosphate dehydrogenase deficiency. Lancet 2008; 371:64–72.

Gladwin MT, Vichinsey E. Pulmonary complications of sickle cell disease. N Engl J Med 2008; 357:2254–2265.

Shander A, Sazama K. Clinical consequences of iron overload from chronic red blood cell transfusions, its diagnosis, and its management by chelation therapy. Transfusion 2010; 50:1144–1155.

Taher AT, Musallam KM, Cappellini MD et al. Optimal management of β thalassaemia intermedia. Br J Haematol 2011; 152:512–523.

Investigations

Blood count is normal between attacks.

During an attack the blood film may show irregularly contracted cells, bite cells (cells with an indentation of the membrane), blister cells (cells in which the Hb appears to have become partially detached from the cell membrane; Fig. 8.26), Heinz bodies (best seen on films stained with methyl violet) and reticulocytosis.

Haemolysis is evident (see p. 387).

G6PD deficiency can be detected using several screening tests, such as demonstration of the decreased ability of G6PD-deficient cells to reduce dyes. The level of the enzyme may also be directly assayed. There are two diagnostic problems. Immediately after an attack the screening tests may be normal (because the oldest red cells with least 6GPD activity are destroyed selectively). The diagnosis of heterozygous females may be difficult because the enzyme level may range from very low to normal depending on lyonization. However, the risk of clinically significant haemolysis is minimal in patients with borderline G6PD activity.

Figure 8.26 ‘Blister’ cells (arrowed) in G6PD deficiency.

Treatment

Any offending drugs should be stopped.

Underlying infection should be treated.

Blood transfusion may be life-saving.

Splenectomy is not usually helpful.

Pyruvate kinase deficiency

This is the most common defect of red cell metabolism after G6PD deficiency; it affects thousands rather than millions of people. The site of the defect is shown in Figure 8.25. There is reduced production of ATP, causing rigid red cells. Homozygotes have haemolytic anaemia and splenomegaly. It is inherited as an autosomal recessive.

Pyrimidine 5′ nucleotidase deficiency

This autosomal disorder produces a haemolytic anaemia with basophilic stippling of the red cells. The enzyme degrades pyrimidine nucleotides to cytidine and uridine (pentose phosphate shunt), which in turn leads to the degradation of RNA in the reticulocytes. Lack of the enzyme results in accumulation of partially degraded RNA, which shows as basophilic stippling in mature red cells. The enzyme is also inhibited by lead (see p. 922) and thus basophilic stippling is seen in lead poisoning. The hereditary form can be diagnosed by measuring the enzyme in erythrocytes. A screening test using the ultraviolet absorption spectrum of red cells is available.

‘Warm’ autoimmune haemolytic anaemias

‘Cold’ autoimmune haemolytic anaemias

Normally, low titres of IgM cold agglutinins reacting at 4°C are present in plasma and are harmless. At low temperatures, these antibodies can attach to red cells and cause their agglutination in the cold peripheries of the body. In addition, activation of complement may cause intravascular haemolysis when the cells return to the higher temperatures in the core of the body.

After certain infections (such as Mycoplasma, cytomegalovirus, Epstein–Barr virus (EBV)), there is increased synthesis of polyclonal cold agglutinins producing a mild to moderate transient haemolysis.

Haemolytic disease of the newborn (HDN)

HDN is due to fetomaternal incompatibility for red cell antigens. Maternal alloantibodies against fetal red cell antigens pass from the maternal circulation via the placenta into the fetus, where they destroy the fetal red cells. Only IgG antibodies are capable of transplacental passage from mother to fetus.

The most common type of HDN is that due to ABO incompatibility, where the mother is usually group O and the fetus group A.

HDN due to ABO incompatibility is usually mild and exchange transfusion is rarely needed. HDN due to RhD incompatibility has become much less common in developed countries following the introduction of anti-D prophylaxis (see below). HDN may be caused by antibodies against antigens in many blood group systems (e.g. other Rh antigens such as c and E, and Kell, Duffy and Kidd; see p. 407).

Sensitization occurs as a result of passage of fetal red cells into the maternal circulation (which most readily occurs at the time of delivery), so that first pregnancies are rarely affected. However, sensitization may occur at other times, for example after a miscarriage, ectopic pregnancy or blood transfusion, or due to episodes during pregnancy which cause transplacental bleeding such as amniocentesis, chorionic villus sampling and threatened miscarriage.

Paroxysmal nocturnal haemoglobinuria (PNH)

Paroxysmal nocturnal haemogloblinuria is a rare form of haemolytic anaemia which results from the clonal expression of haematopoietic stems cells that have mutations in the X-linked gene PIG-A. These mutations result in an impaired synthesis of glycosylphosphatidylinositol (GPI), which anchors many proteins to the cell surface such as decay accelerating factor (DAF; CD55) and membrane inhibitor of reactive lysis (MIRL; CD59) to cell membranes. CD55 and CD59 and other proteins are involved in complement degradation (at the C3 and C5 levels), and in their absence the haemolytic action of complement continues.