Pathogenesis.

The remarkable elasticity and durability of the normal red cell are attributable to the physicochemical properties of its specialized membrane skeleton (Fig. 14-2), which lies closely apposed to the internal surface of the plasma membrane. Its chief protein component, spectrin, consists of two polypeptide chains, α and β, which form intertwined (helical) flexible heterodimers. The “head” regions of spectrin dimers self-associate to form tetramers, while the “tails” associate with actin oligomers. Each actin oligomer can bind multiple spectrin tetramers, thus creating a two-dimensional spectrin-actin skeleton that is connected to the cell membrane by two distinct interactions. The first, involving the proteins ankyrin and band 4.2, binds spectrin to the transmembrane ion transporter, band 3. The second, involving protein 4.1, binds the “tail” of spectrin to another transmembrane protein, glycophorin A.

FIGURE 14-2 Role of the red cell membrane skeleton in hereditary spherocytosis. The left panel shows the normal organization of the major red cell membrane skeletal proteins. Various mutations involving α-spectrin, β-spectrin, ankyrin, band 4.2, or band 3 that weaken the interactions between these proteins cause red cells to lose membrane fragments. To accommodate the resultant change in the ratio of surface area to volume these cells adopt a spherical shape. Spherocytic cells are less deformable than normal ones and therefore become trapped in the splenic cords, where they are phagocytosed by macrophages. GP, glycophorin.

HS is caused by diverse mutations that lead to an insufficiency of membrane skeletal components. As a result of these alterations, the life span of the affected red cells is decreased on average to 10 to 20 days from the normal 120 days. The pathogenic mutations most commonly affect ankyrin, band 3, spectrin, or band 4.2, the proteins involved in the first of the two tethering interactions, presumably because this complex is particularly important in stabilizing the lipid bilayer. Most mutations cause shifts in reading frame or introduce premature stop codons, such that the mutated allele fails to produce any protein. The defective synthesis of the affected protein reduces the assembly of the skeleton as a whole and results in a decrease in the density of the membrane skeleton components. Compound heterozygosity for two defective alleles understandably results in a more severe membrane skeleton deficiency. Young HS red cells are normal in shape, but the deficiency of membrane skeleton reduces the stability of the lipid bilayer, leading to the loss of membrane fragments as red cells age in the circulation. The loss of membrane relative to cyt oplasm “forces” the cells to assume the smallest possible diameter for a given volume, namely, a sphere.

The invariably beneficial effects of splenectomy prove that the spleen has a cardinal role in the premature demise of spherocytes. The travails of spherocytic red cells are fairly well defined. In the life of the portly inflexible spherocyte, the spleen is the villain. Normal red cells must undergo extreme deformation to leave the cords of Billroth and enter the sinusoids. Because of their spheroidal shape and reduced deformability, the hapless spherocytes are trapped in the splenic cords, where they provide a happy meal for phagocytes. The splenic environment also somehow exacerbates the tendency of HS red cells to lose membrane along with K⁺ ions and H₂O; prolonged splenic exposure (erythrostasis), depletion of red cell glucose, and diminished red cell pH have all been suggested to contribute to these abnormalities (Fig. 14-3). After splenectomy the spherocytes persist, but the anemia is corrected.

FIGURE 14-3 Pathophysiology of hereditary spherocytosis.

Morphology. The most specific morphologic finding is spherocytosis, apparent on smears as abnormally small, dark-staining (hyperchromic) red cells lacking the central zone of pallor (Fig. 14-4). Spherocytosis is distinctive but not pathognomonic, since other forms of membrane loss, such as in autoimmune hemolytic anemias, also cause the formation of spherocytes. Other features are common to all hemolytic anemias. These include reticulocytosis, marrow erythroid hyperplasia, hemosiderosis, and mild jaundice. Cholelithiasis (pigment stones) occurs in 40% to 50% of affected adults. Moderate splenic enlargement is characteristic (500–1000 gm); in few other hemolytic anemias is the spleen enlarged as much or as consistently. Splenomegaly results from congestion of the cords of Billroth and increased numbers of phagocytes needed to clear the spherocytes.

FIGURE 14-4 Hereditary spherocytosis (peripheral smear). Note the anisocytosis and several dark-appearing spherocytes with no central pallor. Howell-Jolly bodies (small dark nuclear remnants) are also present in red cells of this asplenic patient.

(Courtesy of Dr. Robert W. McKenna, Department of Pathology, University of Texas Southwestern Medical School, Dallas, TX.)

Clinical Features.

The diagnosis is based on family history, hematologic findings, and laboratory evidence. In two thirds of the patients the red cells are abnormally sensitive to osmotic lysis when incubated in hypotonic salt solutions, which causes the influx of water into spherocytes with little margin for expansion. HS red cells also have an increased mean cell hemoglobin concentration, due to dehydration caused by the loss of K⁺ and H₂O.

The characteristic clinical features are anemia, splenomegaly, and jaundice. The severity of HS varies greatly. In a small minority (mainly compound heterozygotes) HS presents at birth with marked jaundice and requires exchange transfusions. In 20% to 30% of patients the disease is so mild as to be virtually asymptomatic; here the decreased red cell survival is readily compensated for by increased erythropoiesis. In most, however, the compensatory changes are outpaced, producing a chronic hemolytic anemia of mild to moderate severity. The generally stable clinical course is sometimes punctuated by aplastic crises, usually triggered by an acute parvovirus infection. Parvovirus infects and kills red cell progenitors, causing red cell production to cease until an effective immune response commences, generally in 1 to 2 weeks. Because of the reduced life span of HS red cells, cessation of erythropoiesis for even short time periods leads to sudden worsening of the anemia. Transfusions may be necessary to support the patient until the immune response clears the infection. Hemolytic crises are produced by intercurrent events leading to increased splenic destruction of red cells (e.g., infectious mononucleosis); these are clinically less significant than aplastic crises. Gallstones, found in many patients, can also produce symptoms. Splenectomy treats the anemia and its complications, but brings with it the risk of sepsis.

Pathogenesis.

HbS molecules undergo polymerization when deoxygenated. Initially the red cell cytosol converts from a freely flowing liquid to a viscous gel as HbS aggregates form. With continued deoxygenation aggregated HbS molecules assemble into long needle-like fibers within red cells, producing a distorted sickle or holly-leaf shape.

The presence of HbS underlies the major pathologic manifestations: (1) chronic hemolysis, (2) microvascular occlusions, and (3) tissue damage. Several variables affect the rate and degree of sickling:

Sickling causes cumulative damage to red cells through several mechanisms. As HbS polymers grow, they herniate through the membrane skeleton and project from the cell ensheathed by only the lipid bilayer. This severe derangement in membrane structure causes the influx of Ca²⁺ions, which induce the cross-linking of membrane proteins and activate an ion channel that permits the efflux of K⁺ and H₂O. With repeated episodes of sickling, red cells become increasingly dehydrated, dense, and rigid (Fig. 14-7). Eventually, the most severely damaged cells are converted to end-stage, nondeformable, irreversibly sickled cells, which retain a sickle shape even when fully oxygenated. The severity of the hemolysis correlates with the percentage of irreversibly sickled cells, which are rapidly sequestered and removed by mononuclear phagocytes (extravascular hemolysis). Sickled red cells are also mechanically fragile, leading to some intravascular hemolysis as well.

FIGURE 14-7 Pathophysiology of sickle cell disease.

The pathogenesis of the microvascular occlusions, which are responsible for the most serious clinical features, is less certain. Microvascular occlusions are not related to the number of irreversibly sickled cells in the blood, but instead may be dependent upon more subtle red cell membrane damage and other factors, such as inflammation, that tend to slow or arrest the movement of red cells through microvascular beds (see Fig. 14-7). As mentioned above, sickle red cells express higher than normal levels of adhesion molecules and are sticky. Mediators released from granulocytes during inflammatory reactions up-regulate the expression of adhesion molecules on endothelial cells (Chapter 2) and further enhance the tendency for sickle red cells to get arrested during transit throughthe microvasculature. A possible role for inflammatory cells is suggested by observations showing that the leukocyte count correlates with the frequency of pain crises and other measures of tissue damage. The stagnation of red cells within inflamed vascular beds results in extended exposure to low oxygen tension, sickling, and vascular obstruction. Once started, it is easy to envision how a vicious cycle of sickling, obstruction, hypoxia, and more sickling ensues. Depletion of nitric oxide (NO) may also play a part in the vascular occlusions. Free hemoglobin released from lysed sickle red cells can bind and inactivate NO, which is a potent vasodilator and inhibitor of platelet aggregation. Thus, reduced NO increases vascular tone (narrowing vessels) and enhances platelet aggregation, both of which may contribute to red cell stasis, sickling, and (in some instances) thrombosis.

Morphology. In full-blown sickle cell anemia, the peripheral blood demonstrates variable numbers of irreversibly sickled cells, reticulocytosis, and target cells, which result from red cell dehydration (Fig. 14-8). Howell-Jolly bodies (small nuclear remnants) are also present in some red cells due to the asplenia (see below). The bone marrow is hyperplastic as a result of a compensatory erythroid hyperplasia. Expansion of the marrow leads to bone resorption and secondary new bone formation, resulting in prominent cheekbones and changes in the skull that resemble a crew-cut in x-rays. Extramedullary hematopoiesis can also appear. The increased breakdown of hemoglobin can cause pigment gallstones and hyperbilirubinemia.

FIGURE 14-8 Sickle cell disease (peripheral blood smear). A, Low magnification shows sickle cells, anisocytosis, and poikilocytosis. B, Higher magnification shows an irreversibly sickled cell in the center.

(Courtesy of Dr. Robert W. McKenna, Department of Pathology, University of Texas Southwestern Medical School, Dallas, TX.)

In early childhood, the spleen is enlarged up to 500 gm by red pulp congestion, which is caused by the trapping of sickled red cells in the cords and sinuses (Fig. 14-9). With time, however, the chronic erythrostasis leads to splenic infarction, fibrosis, and progressive shrinkage, so that by adolescence or early adulthood only a small nubbin of fibrous splenic tissue is left; this process is called autosplenectomy (Fig. 14-10). Infarctions caused by vascular occlusions can occur in many other tissues as well, including the bones, brain, kidney, liver, retina, and pulmonary vessels, the latter sometimes producing cor pulmonale. In adult patients, vascular stagnation in subcutaneous tissues often leads to leg ulcers; this complication is rare in children.

FIGURE 14-9 A, Spleen in sickle cell disease (low power). Red pulp cords and sinusoids are markedly congested; between the congested areas, pale areas of fibrosis resulting from ischemic damage are evident. B, Under high power, splenic sinusoids are dilated and filled with sickled red cells.

(Courtesy of Dr. Darren Wirthwein, Department of Pathology, University of Texas Southwestern Medical School, Dallas, TX.)

FIGURE 14-10 “Autoinfarcted” splenic remnant in sickle cell disease.

(Courtesy of Dr. Dennis Burns and Dr. Darren Wirthwein, Department of Pathology, University of Texas Southwestern Medical School, Dallas, TX.)

Clinical Features.

Sickle cell disease causes a moderately severe hemolytic anemia (hematocrit 18% to 30%) that is associated with reticulocytosis, hyperbilirubinemia, and the presence of irreversibly sickled cells. Its course is punctuatedby a variety of “crises.” Vaso-occlusive crises, also called pain crises, are episodes of hypoxic injury and infarction that cause severe pain in the affected region. Although infection, dehydration, and acidosis (all of which favor sickling) can act as triggers, in most instances no predisposing cause is identified. The most commonly involved sites are the bones, lungs, liver, brain, spleen, and penis. In children, painful bone crises are extremely common and often difficult to distinguish from acute osteomyelitis. These frequently manifest as the hand-foot syndrome or dactylitis of the bones of the hands or feet, or both. Acute chest syndrome is a particularly dangerous type of vaso-occlusive crisis involving the lungs, which typically presents with fever, cough, chest pain, and pulmonary infiltrates. Pulmonary inflammation (such as may be induced by a simple infection) causes blood flow to become sluggish and “spleenlike,” leading to sickling and vaso-occlusion. This compromises pulmonary function, creating a potentially fatal cycle of worsening pulmonary and systemic hypoxemia, sickling, and vaso-occlusion. Other forms of vascular obstruction, particularly stroke, can take a devastating toll. Factors proposed to contribute to stroke include the adhesion of sickle red cells to arterial vascular endothelium and vasoconstriction caused by the depletion of NO by free hemoglobin.⁸

Although occlusive crises are the most common cause of patient morbidity and mortality, several other acute events complicate the course. Sequestration crises occur in children with intact spleens. Massive entrapment of sickle red cells leads to rapid splenic enlargement, hypovolemia, and sometimes shock. These complications may be fatal in several cases. Survival from sequestration crises and the acute chest syndrome requires treatment with exchange transfusions. Aplastic crises stem from the infection of red cell progenitors by parvovirus B19, which causes a transient cessation of erythropoiesis and a sudden worsening of the anemia.

In addition to these dramatic crises, chronic tissue hypoxia takes a subtle but important toll. Chronic hypoxia is responsible for a generalized impairment of growth and development, as well as organ damage affecting spleen, heart, kidneys, and lungs. Sickling provoked by hypertonicity in the renal medulla causes damage that eventually leads to hyposthenuria (the inability to concentrate urine), which increases the propensity for dehydration and its attendant risks.

Increased susceptibility to infection with encapsulated organisms is another threat. This is due in large part to altered splenic function, which is severely impaired in children by congestion and poor blood flow, and completely absent in adults because of splenic infarction. Defects of uncertain etiology in the alternative complement pathway also impair the opsonization of bacteria. Pneumococcus pneumoniae and Haemophilus influenzae septicemia and meningitis, common causes of death, particularly in children, can be reduced by vaccination and prophylactic antibiotics.

It must be emphasized that there is great variation in the clinical manifestations of sickle cell disease. Some individuals are crippled by repeated vaso-occlusive crises, whereas others have only mild symptoms. The basis for this wide range in disease expression is not understood.

The diagnosis is suggested by the clinical findings and the presence of irreversibly sickled red cells and is confirmed by various tests for sickle hemoglobin. In general, these involve mixing a blood sample with an oxygen-consuming reagent, such as metabisulfite, which induces sickling of red cells if HbS is present. Hemoglobin electrophoresis is also used to demonstrate the presence of HbS and exclude other sickle syndromes, such as HbSC disease. Prenatal diagnosis is possible by analysis of fetal DNA obtained by amniocentesis or chorionic biopsy.

The outlook for patients with sickle cell disease has improved considerably over the last 10 to 20 years. About 90% of patients survive to age 20, and close to 50% survive beyond the fifth decade. A mainstay of treatment is an inhibitor of DNA synthesis, hydroxyurea, which has several beneficial effects. These include (1) an increase in red cell HbF levels, which occurs by unknown mechanisms; and (2) an anti-inflammatory effect, which stems from an inhibition of white cell production. These activities (and possibly others⁹) are believed to act in concert to decrease crises related to vascular occlusions in both children and adults.

Molecular Pathogenesis.

The causative mutations fall into two categories: (1) β⁰ mutations, associated with absent β-globin synthesis, and (2) β⁺ mutations, characterized by reduced (but detectable) β-globin synthesis. Sequencing of β-thalassemia genes has revealed more than 100 different causative mutations, mostly consisting of point mutations. Details of these mutations and their effects are found in specialized texts. Figure 14-11 gives a few illustrative examples.

FIGURE 14-11 Distribution of β-globin gene mutations associated with β-thalassemia. Arrows denote sites where point mutations giving rise to β⁰ or β⁺ thalassemia have been identified.

Impaired β-globin synthesis results in anemia by two mechanisms (Fig. 14-12). The deficit in HbA synthesis produces “underhemoglobinized” hypochromic, microcytic red cells with subnormal oxygen transport capacity. Even more important is the diminished survival of red cells and their precursors, which results from the imbalance in α- and β-globin synthesis. Unpaired α chains precipitate within red cell precursors, forming insoluble inclusions. These inclusions cause a variety of untoward effects, but membrane damage is the proximal cause of most red cell pathology. Many red cell precursors succumb to membrane damage and undergo apoptosis. In severe β-thalassemia, it is estimated that 70% to 85% of red cell precursors suffer this fate, which leads to ineffective erythropoiesis. Those red cells that are released from the marrow also bear inclusions and membrane damage and are prone to splenic sequestration and extravascular hemolysis.

FIGURE 14-12 Pathogenesis of β-thalassemia major. Note that the aggregates of unpaired α-globin chains, a hallmark of the disease, are not visible in routinely stained blood smears. Blood transfusions are a double-edged sword, diminishing the anemia and its attendant complications, but also adding to the systemic iron overload.

In severe β-thalassemia, ineffective erythropoiesis creates several additional problems. Erythropoietic drive in the setting of severe uncompensated anemia leads to massive erythroid hyperplasia in the marrow and extensive extramedullary hematopoiesis. The expanding mass of red cell precursors erodes the bony cortex, impairs bone growth, and produces skeletal abnormalities (described later). Extramedullary hematopoiesis involves the liver, spleen, and lymph nodes, and in extreme cases produces extraosseous masses in the thorax, abdomen, and pelvis. The metabolically active erythroid progenitors steal nutrients from other tissues that are already oxygen-starved, causing severe cachexia in untreated patients.

Another serious complication of ineffective erythropoiesis is the excessive absorption of dietary iron. Ineffective erythropoiesis suppresses the circulating levels of hepcidin, a critical negative regulator of iron absorption (described later under iron deficiency anemia). Low levels of hepcidin and the iron load of repeated blood transfusions inevitably lead to severe iron overload unless preventive steps are taken. Secondary injury to parenchymal organs, particularly the iron-laden liver, often follows and sometimes induces secondary hemochromatosis (Chapter 18).

Clinical Syndromes.

The relationships of clinical phenotypes to underlying genotypes are summarized in Table 14-3. Clinical classification of β-thalassemia is based on the severity of the anemia, which in turn depends on the genetic defect (β⁺ or β⁰) and the gene dosage (homozygous or heterozygous). In general, individuals with two β-thalassemia alleles (β⁺/β⁺, β⁺/β⁰, or β⁰/β⁰) have a severe, transfusion-dependent anemia called β-thalassemia major. Heterozygotes with one βthalassemia gene and one normal gene (β⁺/β or β⁰/β) usually have a mild asymptomatic microcytic anemia. This condition is referred to as β-thalassemia minor or β-thalassemia trait. A third genetically heterogeneous variant of moderate severity is called β-thalassemia intermedia. This category includes milder variants of β⁺/β⁺ or β⁺/β⁰-thalassemia and unusual forms of heterozygous β-thalassemia. Some patients with β-thalassemia intermedia have two defective β-globin genes and an α-thalassemia gene defect, which lessens the imbalance in α- and β-chain synthesis. In other rare but informative cases, individuals have a single β-globin defect and one or two extra copies of normal α-globin genes (stemming from a gene duplication event), which worsens the chain imbalance.¹⁰ These unusual forms of the disease serve to emphasize the cardinal role of unpaired α-globin chains in the pathology. The clinical and morphologic features of β-thalassemia intermedia are not described separately but can be surmised from the following discussions of β-thalassemia major and β-thalassemia minor.

TABLE 14-3 Clinical and Genetic Classification of Thalassemias

β-Thalassemia Major.

β-thalassemia major is most common in Mediterranean countries, parts of Africa, and Southeast Asia. In the United States the incidence is highest in immigrants from these areas. The anemia manifests 6 to 9 months after birth as hemoglobin synthesis switches from HbF to HbA. In untransfused patients, hemoglobin levels are 3 to 6 gm/dL. The red cells may completely lack HbA (β⁰/β⁰ genotype) or contain small amounts (β⁺/β⁺ or β⁰/β⁺ genotypes). The major red cell hemoglobin is HbF, which is markedly elevated. HbA₂ levels are sometimes high but more often are normal or low.

Morphology. Blood smears show severe red cell abnormalities, including marked variation in size (anisocytosis) and shape (poikilocytosis), microcytosis, and hypochromia. Target cells (so called because hemoglobin collects in the center of the cell), basophilic stippling, and fragmented red cells are also common. Inclusions of aggregated α chains are efficiently removed by the spleen and not easily seen. The reticulocyte count is elevated, but it is lower than expected for the severity of anemia because of the ineffective erythropoiesis. Variable numbers of poorly hemoglobinized nucleated red cell precursors (normoblasts) are seen in the peripheral blood as a result of “stress” erythropoiesis and abnormal release from sites of extramedullary hematopoiesis.

Other major alterations involve the bone marrow and spleen. In the untransfused patient there is a striking expansion of hematopoietically active marrow. In the bones of the face and skull the burgeoning marrow erodes existing cortical bone and induces new bone formation, giving rise to a “crew-cut” appearance on x-ray (Fig. 14-13). Both phagocyte hyperplasia and extramedullary hematopoiesis contribute to enlargement of the spleen, which can weigh as much as 1500 gm. The liver and the lymph nodes can also be enlarged by extramedullary hematopoiesis.

FIGURE 14-13 Thalassemia: x-ray film of the skull showing new bone formation on the outer table, producing perpendicular radiations resembling a crewcut.

(Courtesy of Dr. Jack Reynolds, Department of Radiology, University of Texas Southwestern Medical School, Dallas, TX.)

Hemosiderosis and secondary hemochromatosis, the two manifestations of iron overload (Chapter 18), occur in almost all patients. The deposited iron often damages organs, most notably the heart, liver, and pancreas.

The clinical course of β-thalassemia major is brief unless blood transfusions are given. Untreated children suffer from growth retardation and die at an early age from the effects of anemia. In those who survive long enough, the cheekbones and other bony prominences are enlarged and distorted. Hepatosplenomegaly due to extramedullary hematopoiesis is usually present. Although blood transfusions improve the anemia and suppress complications related to excessive erythropoiesis, they lead to complications of their own. Cardiac disease resulting from progressive iron overload and secondary hemochromatosis (Chapter 18) is an important cause of death, particularly in heavily transfused patients, who must be treated with iron chelators to prevent or reduce this complication. With transfusions and iron chelation, survival into the third decade is possible, but the overall outlook remains guarded. Bone marrow transplantation is the only therapy offering a cure and is being used increasingly.¹¹ Prenatal diagnosis is possible by molecular analysis of DNA.

Clinical Syndromes.

The clinical syndromes are determined and classified by the number of α-globin genes that are deleted. Each of the four α-globin genes normally contributes 25% of the total α-globin chains. α-Thalassemia syndromes stem from combinations of deletions that remove one to four α-globin genes. Not surprisingly, the severity of the clinical syndrome is proportional to the number of α-globin genes that are deleted. The different types of α-thalassemia and their salient clinical features are listed in Table 14-3.

Warm Antibody Type.

This is the most common form of immunohemolytic anemia. About 50% of cases are idiopathic (primary); the others are related to a predisposing condition (see Table 14-4) or exposure to a drug. Most causative antibodies are of the IgG class; less commonly, IgA antibodies are culpable. The red cell hemolysis is mostly extravascular. IgG-coated red cells bind to Fc receptors on phagocytes, which remove red cell membrane during “partial” phagocytosis. As in hereditary spherocytosis, the loss of membrane converts the red cells to spherocytes, which are sequestered and removed in the spleen. Moderate splenomegaly due to hyperplasia of splenic phagocytes is usually seen.

As with other autoimmune disorders, the cause of primary immunohemolytic anemia is unknown. In many cases, the antibodies are directed against the Rh blood group antigens. The mechanisms of drug-induced immunohemolytic anemia are better understood. Two different mechanisms have been described.

Cold Agglutinin Type.

This form of immunohemolytic anemia is caused by IgM antibodies that bind red cells avidly at low temperatures (0°–4°C).¹⁴ It is less common than warm antibody immunohemolytic anemia, accounting for 15% to 30% of cases. Cold agglutinin antibodies sometimes appear transiently following certain infections, such as with Mycoplasma pneumoniae, Epstein-Barr virus, cytomegalovirus, influenza virus, and human immunodeficiency virus (HIV). In these settings the disorder is self-limited and the antibodies rarely induce clinically important hemolysis. Chronic cold agglutinin immunohemolytic anemia occurs in association with certain B-cell neoplasms or as an idiopathic condition.

Clinical symptoms result from binding of IgM to red cells in vascular beds where the temperature may fall below 30°C, such as in exposed fingers, toes, and ears. IgM binding agglutinates red cells and fixes complement rapidly. As the blood recirculates and warms, IgM is released, usually before complement-mediated hemolysis can occur. However, the transient interaction with IgM is sufficient to deposit sublytic quantities of C3b, an excellent opsonin, which leads to the removal of affected red cells by phagocytes in the spleen, liver, and bone marrow. The hemolysis is of variable severity. Vascular obstruction caused by agglutinated red cells results in pallor, cyanosis, and Raynaud phenomenon (Chapter 11) in body parts exposed to cold temperature.

Cold Hemolysin Type.

Cold hemolysins are autoantibodies responsible for an unusual entity known as paroxysmal cold hemoglobinuria. This rare disorder causes substantial, sometimes fatal, intravascular hemolysis and hemoglobinuria. The autoantibodies are IgGs that bind to the P blood group antigen on the red cell surface¹⁴ in cool, peripheral regions of the body. Complement-mediated lysis occurs when the cells recirculate to warm central regions, since the complement cascade functions more efficiently at 37°C. Most cases are seen in children following viral infections; in this setting the disorder is transient, and most of those affected recover within 1 month.

Treatment of warm antibody immunohemolytic anemia centers on the removal of initiating factors (i.e., drugs); when this is not feasible, immunosuppressive drugs and splenectomy are the mainstays.¹⁵ Chronic cold agglutinin immunohemolytic anemia caused by IgM antibodies is more difficult to treat.¹⁴

Normal Vitamin B₁₂Metabolism.

Vitamin B₁₂ is a complex organometallic compound known as cobalamin. Under normal circumstances humans are totally dependent on dietary vitamin B₁₂. Microorganisms are the ultimate origin of cobalamin in the food chain. Plants and vegetables contain little cobalamin, save that contributed by microbial contamination, and strictly vegetarian or macrobiotic diets do not provide adequate amounts of this essential nutrient. The daily requirement is 2 to 3 μg. A diet that includes animal products contains significantly larger amounts and normally results in the accumulation of intrahepatic stores of vitamin B₁₂ that are sufficient to last for several years.

Absorption of vitamin B₁₂ requires intrinsic factor, which is secreted by the parietal cells of the fundic mucosa (Fig. 14-18). Vitamin B₁₂ is freed from binding proteins in food through the action of pepsin in the stomach and binds to salivary proteins called cobalophilins, or R-binders. In the duodenum, bound vitamin B₁₂ is released by the action of pancreatic proteases. It then associates with intrinsic factor. This complex is transported to the ileum, where it is endocytosed by ileal enterocytes that express intrinsic factor receptors on their surfaces. Within ileal cells, vitamin B₁₂ associates with a major carrier protein, transcobalamin II, and is secreted into the plasma. Transcobalamin II delivers vitamin B₁₂ to the liver and other cells of the body, including rapidly proliferating cells in the bone marrow and the gastrointestinal tract. In addition to this major pathway, there is also a poorly understood alternative uptake mechanism that is not dependent on intrinsic factor or an intact terminal ileum. Up to 1% of a large oral dose can be absorbed by this pathway, making it feasible to treat pernicious anemia with high doses of oral vitamin B₁₂.

FIGURE 14-18 Schematic illustration of vitamin B₁₂ absorption. IF, intrinsic factor; R-binders, cobalophilins (see text).

Biochemical Functions of Vitamin B₂.

Only two reactions in humans are known to require vitamin B₁₂. In one, methylcobalamin serves as an essential cofactor in the conversion of homocysteine to methionine by methionine synthase (Fig. 14-19). In the process, methylcobalamin yields a methyl group that is recovered from N⁵-methyltetrahydrofolic acid (N⁵-methyl FH₄), the principal form of folic acid in plasma. In the same reaction, N⁵-methyl FH₄ is converted to tetrahydrofolic acid (FH₄). FH₄ is crucial, since it is required (through its derivative N^5,10-methylene FH₄) for the conversion of deoxyuridine monophosphate (dUMP) to deoxythymidine monophosphate (dTMP), an immediate precursor of DNA. It is postulated that the fundamental cause of the impaired DNA synthesis in vitamin B₁₂ deficiency is the reduced availability of FH₄, most of which is “trapped” as N⁵-methyl FH₄. The FH₄ deficit may be further exacerbated by an “internal” folate deficiency caused by a failure to synthesize metabolically active polyglutamylated forms. This stems from the requirement for vitamin B₁₂ in the synthesis of methionine, which contributes a carbon group needed in the metabolic reactions that create folate polyglutamates (Fig. 14-20). Whatever the mechanism, lack of folate is the proximate cause of anemia in vitamin B₁₂ deficiency, since the anemia improves with administration of folic acid.

FIGURE 14-19 Relationship of N⁵-methyl FH₄, methionine synthase, and thymidylate synthetase. In cobalamin (Cbl) deficiency, folate is sequestered as N⁵-methyl FH₄. This ultimately deprives thymidylate synthetase of its folate coenzyme (N^5,10-methylene FH₄), thereby impairing DNA synthesis. FH₄, tetrahydrofolic acid.

FIGURE 14-20 Role of folate derivatives in the transfer of one-carbon fragments for synthesis of biologic macromolecules. FH₄, tetrahydrofolic acid; FH₂, dihydrofolic acid; FIGlu, formiminoglutamate; dTMP, deoxythymidine monophosphate.

The neurologic complications associated with vitamin B₁₂ deficiency are more enigmatic, since they are not improved by folate administration. The other known reaction that depends on vitamin B₁₂ is the isomerization of methylmalonyl coenzyme A to succinyl coenzyme A, which requires adenosylcobalamin as a prosthetic group on the enzyme methylmalonyl–coenzyme A mutase. A deficiency of vitamin B₁₂ thus leads to increased plasma and urine levels of methylmalonic acid. Interruption of this reaction and the consequent buildup of methylmalonate and propionate (a precursor) could lead to the formation and incorporation of abnormal fatty acids into neuronal lipids. It has been suggested that this biochemical abnormality predisposes to myelin breakdown and thereby produces the neurologic complications of vitamin B₁₂ deficiency (Chapter 28). However, rare individuals with hereditary deficiencies of methylmalonyl–coenzyme A mutase, while having complications related to methylmalonyl acidemia, do not suffer from the neurologic abnormalities seen in vitamin B₁₂ deficiency, casting doubt on this explanation.

Having completed our overview of vitamin B₁₂ metabolism, we can now turn to pernicious anemia.

Incidence.

Although somewhat more prevalent in Scandinavian and other Caucasian populations, pernicious anemia occurs in all racial groups, including blacks and Hispanics. It is a disease of older adults; the median age at diagnosis is 60 years, and it is rare in people younger than 30. A genetic predisposition is strongly suspected, but no definable genetic pattern of transmission has been discerned. As described below, many affected individuals have a tendency to form antibodies against multiple self-antigens.

Pathogenesis.

Pernicious anemia is believed to result from an autoimmune attack on the gastric mucosa. Histologically, there is a chronic atrophic gastritis marked by a loss of parietal cells, a prominent infiltrate of lymphocytes and plasma cells, and megaloblastic changes in mucosal cells similar to those found in erythroid precursors. Three types of autoantibodies are present in many, but not all, patients. About 75% of patients have a type I antibody that blocks the binding of vitamin B₁₂ to intrinsic factor. Type I antibodies are found in both plasma and gastric juice. Type II antibodies prevent binding of the intrinsic factor–vitamin B₁₂ complex to its ileal receptor. These antibodies are also found in a large proportion of patients with pernicious anemia. Type III antibodies, present in 85% to 90% of patients, recognize the α and β subunits of the gastric proton pump, which is normally localized to the microvilli of the canalicular system of the gastric parietal cell. These antibodies are not specific for pernicious anemia or other autoimmune diseases, since they are found in as many as 50% of elderly persons with idiopathic chronic gastritis not associated with pernicious anemia.

Autoantibodies are of diagnostic utility, but they are not thought to be the primary cause of the gastric pathology. Rather, it seems that an autoreactive T-cell response initiates gastric mucosal injury and triggers the formation of autoantibodies, which may exacerbate the epithelial injury. When the mass of intrinsic factor–secreting cells falls below a threshold (and reserves of stored vitamin B₁₂ are depleted), anemia develops. In an animal model of autoimmune gastritis mediated by CD4+ T cells, a pattern of autoantibodies resembling that seen in pernicious anemia develops, thus supporting the primacy of T-cell autoimmunity. The common association of pernicious anemia with other autoimmune disorders, particularly autoimmune thyroiditis and adrenalitis, is also consistent with an underlying immune basis. The tendency to develop multiple autoimmune disorders, including pernicious anemia, is linked to specific sequence variants of NALP1,¹⁶ an innate immune receptor that maps to chromosome 17p13.

Vitamin B₁₂ deficiency is associated with disorders other than pernicious anemia. Most of these impair absorption of the vitamin at one of the steps outlined earlier (see Table 14-5). With achlorhydria and loss of pepsin secretion (which occurs in some elderly individuals), vitamin B₁₂ is not readily released from proteins in food. With gastrectomy, intrinsic factor is not available for uptake in the ileum. With loss of exocrine pancreatic function, vitamin B₁₂ cannot be released from R-binder–vitamin B₁₂ complexes. Ileal resection or diffuse ileal disease can remove or damage the site of intrinsic factor–vitamin B₁₂ complex absorption. Tapeworms compete with the host for B₁₂ and can induce a deficiency state. In some settings, such as pregnancy, hyperthyroidism, disseminated cancer, and chronic infection, an increased demand for vitamin B₁₂ can produce a relative deficiency, even with normal absorption.

Clinical Features.

Pernicious anemia is insidious in onset, so the anemia is often quite severe by the time the affected person seeks medical attention. The course is progressive unless halted by therapy.

The diagnosis is based on (1) a moderate to severe megaloblastic anemia, (2) leukopenia with hypersegmented granulocytes, (3) low serum vitamin B₁₂, and (4) elevated levels of homocysteine and methylmalonic acid in the serum. The diagnosis is confirmed by a striking increase in reticulocytes and an improvement in hematocrit levels beginning about 5 days after parenteral administration of vitamin B₁₂. Serum antibodies to intrinsic factor are highly specific for pernicious anemia. Their presence attests to the cause rather than the presence or absence of vitamin B₁₂ deficiency.

Persons with atrophic and metaplastic changes in the gastric mucosa associated with pernicious anemia are at increased risk of developing gastric carcinoma (Chapter 17). As mentioned, serum homocysteine levels are raised in individuals with vitamin B₁₂ deficiency. Elevated homocysteine levels are a risk factor for atherosclerosis and thrombosis, and it is suspected that vitamin B₁₂ deficiency may increase the incidence of vascular disease. With parenteral or high-dose oral vitamin B₁₂, the anemia can be cured and the peripheral neurologic changes reversed or at least halted in their progression, but the changes in the gastric mucosa and the risk of carcinoma are unaffected.

Etiology.

The three major causes of folic acid deficiency are (1) decreased intake, (2) increased requirements, and (3) impaired utilization (see Table 14-5). Humans are entirely dependent on dietary sources for their folic acid requirement, which is 50 to 200 μg daily. Most normal diets contain ample amounts. The richest sources are green vegetables such as lettuce, spinach, asparagus, and broccoli. Certain fruits (e.g., lemons, bananas, melons) and animal sources (e.g., liver) contain lesser amounts. The folic acid in these foods is largely in the form of folylpolyglutamates. Despite their abundance in raw foods, polyglutamates are sensitive to heat; boiling, steaming, or frying of foods for 5 to 10 minutes destroys up to 95% of the folate content. Intestinal conjugases split the polyglutamates into monoglutamates that are readily absorbed in the proximal jejunum. During intestinal absorption they are modified to 5-methyltetrahydrofolate, the normal transport form of folate. The body’s reserves of folate are relatively modest, and a deficiency can arise within weeks to months if intake is inadequate.

Decreased intake can result from either a nutritionally inadequate diet or impairment of intestinal absorption. A normal diet contains folate in excess of the minimal daily adult requirement. Inadequate dietary intakes are almost invariably associated with grossly deficient diets. Such dietary inadequacies are most frequently encountered in chronic alcoholics, the indigent, and the very elderly. In alcoholics with cirrhosis, other mechanisms of folate deficiency such as trapping of folate within the liver, excessive urinary loss, and disordered folate metabolism have also been implicated. Under these circumstances, the megaloblastic anemia is often accompanied by general malnutrition and manifestations of other avitaminoses, including cheilosis, glossitis, and dermatitis. Malabsorption syndromes, such as sprue, can lead to inadequate absorption of this nutrient, as can diffuse infiltrative diseases of the small intestine (e.g., lymphoma). In addition, certain drugs, particularly the anticonvulsant phenytoin and oral contraceptives, interfere with absorption.

Despite normal intake of folic acid, a relative deficiency can be encountered when requirements are increased. Conditions in which this is seen include pregnancy, infancy, hematologic derangements associated with hyperactive hematopoiesis (hemolytic anemias), and disseminated cancer. In all these circumstances the demands of increased DNA synthesis render normal intake inadequate.

Folic acid antagonists, such as methotrexate, inhibit dihydrofolate reductase and lead to a deficiency of FH₄. With inhibition of folate metabolism, all rapidly growing cells are affected, but particularly the cells of the bone marrow and the gastrointestinal tract. Many chemotherapeutic drugs used in the treatment of cancer damage DNA or inhibit DNA synthesis through other mechanisms; these can also cause megaloblastic changes in rapidly dividing cells.

As mentioned at the outset, the megaloblastic anemia that results from a deficiency of folic acid is identical to that encountered in vitamin B₁₂ deficiency. Thus, the diagnosis of folate deficiency can be made only by demonstration of decreased folate levels in the serum or red cells. As in vitamin B₁₂ deficiency, serum homocysteine levels are increased, but methylmalonate concentrations are normal. However, neurologic changes do not occur.

Although prompt hematologic response heralded by reticulocytosis follows the administration of folic acid, it should be remembered that the hematologic symptoms of vitamin B₁₂ deficiency anemia also respond to folate therapy. However, folate does not prevent (and may even exacerbate) the neurologic deficits typical of the vitamin B₁₂ deficiency states. It is thus essential to exclude vitamin B₁₂ deficiency in megaloblastic anemia before initiating therapy with folate.

Iron Metabolism.

The normal daily Western diet contains about 10 to 20 mg of iron, most in the form of heme contained in animal products, with the remainder being inorganic iron in vegetables. About 20% of heme iron (in contrast to 1% to 2% of nonheme iron) is absorbable, so the average Western diet contains sufficient iron to balance fixed daily losses. The total body iron content is normally about 2 gm in women and as high as 6 gm in men, and can be divided into functional and storage compartments (Table 14-6). About 80% of the functional iron is found in hemoglobin; myoglobin and iron-containing enzymes such as catalase and the cytochromes contain the rest. The storage pool represented by hemosiderin and ferritin contains about 15% to 20% of total body iron. Healthy young females have smaller stores of iron than do males, primarily because of blood loss during menstruation, and often develop iron deficiency due to excessive losses or increased demands associated with menstruation and pregnancy, respectively.

TABLE 14-6 Iron Distribution in Healthy Young Adults (mg)

Iron in the body is recycled extensively between the functional and storage pools (Fig. 14-21). It is transported in plasma by an iron-binding glycoprotein called transferrin, which is synthesized in the liver. In normal individuals, transferrin is about one third saturated with iron, yielding serum iron levels that average 120 μg/dL in men and 100 μg/dL in women. The major function of plasma transferrin is to deliver iron to cells, including erythroid precursors, which require iron to synthesize hemoglobin. Erythroid precursors possess high-affinity receptors for transferrin, which mediate iron import through receptor-mediated endocytosis.

FIGURE 14-21 Iron metabolism. Iron absorbed from the gut is bound to plasma transferrin and transported to the marrow, where it is delivered to developing red cells and incorporated into hemoglobin. Mature red cells are released into the circulation and, after 120 days, are ingested by macrophages, primarily in the spleen, liver, and bone marrow. Here iron is extracted from hemoglobin and recycled to plasma transferrin. At equilibrium, iron absorbed from the gut is balanced by losses in shed keratinocytes, enterocytes, and (in women) endometrium.

Free iron is highly toxic (as described in Chapter 18), and it is therefore important that storage iron be sequestered. This is achieved by binding iron in the storage pool tightly to either ferritin or hemosiderin. Ferritin is a ubiquitous protein-iron complex that is found at highest levels in the liver, spleen, bone marrow, and skeletal muscles. In the liver, most ferritin is stored within the parenchymal cells; in other tissues, such as the spleen and the bone marrow, it is found mainly in macrophages. Hepatocyte iron is derived from plasma transferrin, whereas storage iron in macrophages is derived from the breakdown of red cells. Intracellular ferritin is located in the cytosol and in lysosomes, in which partially degraded protein shells of ferritin aggregate into hemosiderin granules. Iron in hemosiderin is chemically reactive and turns blue-black when exposed to potassium ferrocyanide, which is the basis for the Prussian blue stain. With normal iron stores, only trace amounts of hemosiderin are found in the body, principally in macrophages in the bone marrow, spleen, and liver. In iron-overloaded cells, most iron is stored in hemosiderin.

Since plasma ferritin is derived largely from the storage pool of body iron, its levels correlate well with body iron stores. In iron deficiency, serum ferritin is always below 12 μg/L, whereas in iron overload values approaching 5000 μg/L can be seen. Of physiologic importance, the storage iron pool can be readily mobilized if iron requirements increase, as may occur after loss of blood.

Iron balance is maintained largely by regulating the absorption of dietary iron in the proximal duodenum. Iron is both essential for cellular metabolism and highly toxic in excess, and total body iron stores must therefore be regulated meticulously. There is no regulated pathway for iron excretion, which is limited to the 1 to 2 mg lost each day through the shedding of mucosal and skin epithelial cells. In contrast, as body iron stores rise, absorption falls, and vice versa. The pathways responsible for the absorption of iron are now understood in reasonable detail (Fig. 14-22), and differ partially for nonheme and heme iron.¹⁷ Luminal nonheme iron is mostly in the Fe³⁺ (ferric) state and must first be reduced to Fe²⁺ (ferrous) iron by ferrireductases, such as b cytochromes and STEAP3. Fe²⁺ iron is then transported across the apical membrane by divalent metal transporter 1 (DMT1). The absorption of nonheme iron is variable and often inefficient, being inhibited by substances in the diet that bind and stabilize Fe³⁺ iron and enhanced by substances that stabilize Fe²⁺ iron (described below). Frequently, less than 5% of dietary nonheme iron is absorbed. In contrast, about 25% of the heme iron derived from hemoglobin, myoglobin, and other animal proteins is absorbed. Heme iron is moved across the apical membrane into the cytoplasm through transporters that are incompletely characterized. Here, it is metabolized to release Fe²⁺ iron, which enters a common pool with nonheme Fe²⁺ iron. Iron that enters the duodenal cells can follow one of two pathways: transport to the blood or storage as mucosal iron. This distribution is influenced by body iron stores, as we shall discuss below. Fe²⁺ iron destined for the circulation, is transported from the cytoplasm across the basolateral enterocyte membrane by ferriportin. This process is coupled to the oxidation of Fe²⁺ iron to Fe³⁺ iron, which is carried out by the iron oxidases hephaestin and ceruloplasmin. Newly absorbed Fe³⁺ iron binds rapidly to the plasma protein transferrin, which delivers iron to red cell progenitors in the marrow (see Fig. 14-21). Both DMT1 and ferriportin are widely distributed in the body and are involved in iron transport in other tissues as well. For example, DMT1 also mediates the uptake of “functional” iron (derived from endocytosed transferrin) across lysosomal membranes into the cytosol of red cell precursors in the bone marrow, and ferriportin plays an important role in the release of storage iron from macrophages.

FIGURE 14-22 Regulation of iron absorption. Duodenal epithelial cell uptake of heme and nonheme iron is depicted. When the storage sites of the body are replete with iron and erythropoietic activity is normal, plasma hepcidin levels are high. This leads to downregulation of ferriportin and trapping of most of the absorbed iron, which is lost when duodenal epithelial cells are shed into the gut. Conversely, when body iron stores decrease or when erythropoiesis is stimulated, hepcidin levels fall and ferriportin activity increases, allowing a greater fraction of the absorbed iron to be transferred to plasma transferrin. DMT1, divalent metal transporter 1.

Iron absorption is regulated by hepcidin, a small circulating peptide that is synthesized and released from the liver in response to increases in intrahepatic iron levels.¹⁷ Hepcidin inhibits iron transfer from the enterocyte to plasma by binding to ferriportin and causing it to be endocytosed and degraded. As a result, as hepcidin levels rise, iron becomes trapped within duodenal cells in the form of mucosal ferritin and is lost as these cells are sloughed. Thus, when the body is replete with iron, high hepcidin levels inhibit its absorption into the blood. Conversely, with low body stores of iron, hepcidin synthesis falls and this in turn facilitates iron absorption. By inhibiting ferriportin, hepcidin not only reduces iron uptake from enetrocytes but also suppresses iron release from macrophages, which are an important source of the iron that is used by erythroid precursors to make hemoglobin. This, as we shall see, is important in the pathogenesis of anemia of chronic diseases.

Alterations in hepcidin have a central role in diseases involving disturbances of iron metabolism. As will be described subsequently, the anemia of chronic disease is caused in part by inflammatory mediators that increase hepatic hepcidin production.¹⁸ A rare form of microcytic anemia is caused by mutations that disable TMPRSS6, a hepatic transmembrane serine protease that normally suppresses hepcidin production when iron stores are low. Affected patients have high hepcidin levels, resulting in reduced iron absorption and failure to respond to iron therapy. Conversely, hepcidin activity is inappropriately low in both primary and secondary hemochromatosis, a syndrome caused by systemic iron overload. Secondary hemochromatosis can occur in diseases associated with ineffective erythropoiesis, such as β-thalassemia major and myelodysplastic syndromes (Chapter 13). Through incompletely understood mechanisms, ineffective erythropoiesis suppresses hepatic hepcidin production, even when iron stores are high. As discussed in Chapter 18, the various inherited forms of primary hematochromatosis are associated with mutations in hepcidin or the genes that regulate hepcidin expression.

Etiology.

Iron deficiency can result from (1) dietary lack, (2) impaired absorption, (3) increased requirement, or (most importantly) (4) chronic blood loss. To maintain a normal iron balance, about 1 mg of iron must be absorbed from the diet every day. Because only 10% to 15% of ingested iron is absorbed, the daily iron requirement is 7 to 10 mg for adult men and 7 to 20 mg for adult women. Since the average daily dietary intake of iron in the Western world is about 15 to 20 mg, most men ingest more than adequate iron, whereas many women consume marginally adequate amounts of iron. The bioavailability of dietary iron is as important as the overall content. Heme iron is much more absorbable than inorganic iron, the absorption of which is influenced by other dietary contents. Absorption of inorganic iron is enhanced by ascorbic acid, citric acid, amino acids, and sugars in the diet, and inhibited by tannates (found in tea), carbonates, oxalates, and phosphates.

Dietary lack is rare in developed countries, where on average about two thirds of the dietary iron is in the readily absorbed heme form provided by meat. The situation is different in developing countries, where food is less abundant and most iron in the diet is found in plants in the poorly absorbable inorganic form. Dietary iron inadequacy occurs in even privileged societies in the following groups:

Impaired absorption is found in sprue, other causes of fat malabsorption (steatorrhea), and chronic diarrhea. Gastrectomy impairs iron absorption by decreasing hydrochloric acid and transit time through the duodenum. Specific items in the diet, as is evident from the preceding discussion, can also affect absorption.

Increased requirement is an important cause of iron deficiency in growing infants, children, and adolescents, as well as premenopausal women, particularly during pregnancy. Economically deprived women having multiple, closely spaced pregnancies are at exceptionally high risk.

Chronic blood loss is the most common cause of iron deficiency in the Western world. External hemorrhage or bleeding into the gastrointestinal, urinary, or genital tracts depletes iron reserves. Iron deficiency in adult men and postmenopausal women in the Western world must be attributed to gastrointestinal blood loss until proven otherwise. To prematurely ascribe iron deficiency in such individuals to any other cause is to run the risk of missing an occult gastrointestinal cancer or other bleeding lesion. An alert clinician investigating unexplained iron deficiency anemia occasionally discovers an occult bleed or cancer and thereby saves a life.

Pathogenesis.

Whatever its basis, iron deficiency produces a hypochromic microcytic anemia. At the outset of chronic blood loss or other states of negative iron balance, reserves in the form of ferritin and hemosiderin may be adequate to maintain normal hemoglobin and hematocrit levels as well as normal serum iron and transferrin saturation. Progressive depletion of these reserves first lowers serum iron and transferrin saturation levels without producing anemia. In this early stage there is increased erythroid activity in the bone marrow. Anemia appears only when iron stores are completely depleted and is accompanied by low serum iron, ferritin, and transferrin saturation levels.

Morphology. The bone marrow reveals a mild to moderate increase in erythroid progenitors. A diagnostically significant finding is the disappearance of stainable iron from macrophages in the bone marrow, which is best assessed by performing Prussian blue stains on smears of aspirated marrow. In peripheral blood smears, the red cells are small (microcytic) and pale (hypochromic). Normal red cells with sufficient hemoglobin have a zone of central pallor measuring about one third of the cell diameter. In established iron deficiency the zone of pallor is enlarged; hemoglobin may be seen only in a narrow peripheral rim (Fig. 14-23). Poikilocytosis in the form of small, elongated red cells (pencil cells) is also characteristically seen.

FIGURE 14-23 Hypochromic microcytic anemia of iron deficiency (peripheral blood smear). Note the small red cells containing a narrow rim of peripheral hemoglobin. Scattered fully hemoglobinized cells, present due to recent blood transfusion, stand in contrast.

(Courtesy of Dr. Robert W. McKenna, Department of Pathology, University of Texas Southwestern Medical School, Dallas, TX.)

Clinical Features.

The clinical manifestations of the anemia are nonspecific and were detailed earlier. The dominating signs and symptoms frequently relate to the underlying cause of the anemia, for example, gastrointestinal or gynecologic disease, malnutrition, pregnancy, and malabsorption. In severe and long-standing iron deficiency, depletion of iron-containing enzymes in cells throughout the body also causes other changes, including koilonychia, alopecia, atrophic changes in the tongue and gastric mucosa, and intestinal malabsorption. Depletion of iron from the central nervous system may lead to the appearance of pica, in which affected individuals consume non-foodstuffs such as clay or food ingredients such as flour, and periodically move their limbs during sleep. Esophageal webs appear together with microcytic hypochromic anemia and atrophic glossitis to complete the triad of major findings in the rare Plummer-Vinson syndrome (Chapter 17).

The diagnosis of iron deficiency anemia ultimately rests on laboratory studies. Both the hemoglobin and hematocrit are depressed, usually to a moderate degree, in association with hypochromia, microcytosis, and modest poikilocytosis. The serum iron and ferritin are low, and the total plasma iron-binding capacity (reflecting elevated transferrin levels) is high. Low serum iron with increased iron-binding capacity results in a reduction of transferrin saturation to below 15%. Reduced iron stores inhibit hepcidin synthesis, and its serum levels fall. In uncomplicated iron deficiency, oral iron supplementation produces an increase in reticulocytes in about 5 to 7 days that is followed by a steady increase in blood counts and the normalization of red cell indices.

Etiology.

The most common circumstances associated with aplastic anemia are listed in Table 14-7. Most cases of “known” etiology follow exposure to chemicals and drugs. Certain drugs and agents (including many cancer chemotherapy drugs and the organic solvent benzene) cause marrow suppression that is dose related and reversible. In other instances, aplastic anemia arises in an unpredictable, idiosyncratic fashion following exposure to drugs that normally cause little or no marrow suppression. The implicated drugs include chloramphenicol and gold salts.

TABLE 14-7 Major Causes of Aplastic Anemia

Persistent marrow aplasia can also appear after a variety of viral infections, most commonly viral hepatitis of the non-A, non-B, non-C, non-G type, which is associated with 5% to 10% of cases. Why aplastic anemia develops in certain individuals is not understood.

Whole-body irradiation can destroy hematopoietic stem cells in a dose-dependent fashion. Persons who receive therapeutic irradiation or are exposed to radiation in nuclear accidents (e.g., Chernobyl) are at risk for marrow aplasia.

Inherited defects underlie some forms of aplastic aplasia. Fanconi anemia is a rare autosomal recessive disorder caused by defects in a multiprotein complex that is required for DNA repair (Chapter 7).²¹ Marrow hypofunction becomes evident early in life and is often accompanied by multiple congenital anomalies, such as hypoplasia of the kidney and spleen and bone anomalies, which most commonly involve the thumbs or radii. Inherited defects in telomerase are found in 5% to 10% of adult-onset aplastic anemia.²² You will recall from Chapters 1 and 7 that telomerase is required for cellular immortality and limitless replication. It might be anticipated, therefore, that partial deficits in telomerase activity could result in premature hematopoietic stem cell exhaustion and marrow aplasia. Even more common than telomerase mutations are abnormally short telomeres, which are found in the marrow cells of as many as half of those affected with aplastic anemia. It is unknown whether this shortening is due to other unappreciated telomerase defects or is a consequence of excessive stem cell replication.

In most instances, however, no initiating factor can be identified; about 65% of cases fall into this idiopathic category.

Pathogenesis.

The pathogenesis of aplastic anemia is not fully understood. Indeed, it is unlikely that a single mechanism underlies all cases. However, two major etiologies have been invoked: an extrinsic, immune-mediated suppression of marrow progenitors; and an intrinsic abnormality of stem cells (Fig. 14-24).

FIGURE 14-24 Pathophysiology of aplastic anemia. Damaged stem cells can produce progeny expressing neoantigens that evoke an autoimmune reaction, or give rise to a clonal population with reduced proliferative capacity. Either pathway could lead to marrow aplasia. See text for abbreviations.

Experimental studies have increasingly focused on a model in which activated T cells suppress hematopoietic stem cells. Stem cells may first be antigenically altered by exposure to drugs, infectious agents, or other unidentified environmental insults. This provokes a cellular immune response, during which activated T_H1 cells produce cytokines such as interferon-γ (IFNγ) and TNF that suppress and kill hematopoietic progenitors. This scenario is supported by several observations. Expression analysis of the few remaining marrow stem cells from aplastic anemia marrows has revealed that genes involved in apoptosis and death pathways are up-regulated; of note, the same genes are up-regulated in normal stem cells exposed to interferon-γ. Even more compelling (and clinically relevant) evidence comes from experience with immunosuppressive therapy. Antithymocyte globulin and other immunosuppressive drugs such as cyclosporine produce responses in 60% to 70% of patients. It is proposed that these therapies work by suppressing or killing autoreactive T-cell clones. The antigens recognized by the autoreactive T cells are not well defined. In some instances GPI-linked proteins may be the targets, possibly explaining the previously noted association of aplastic anemia and PNH.

Alternatively, the notion that aplastic anemia results from a fundamental stem cell abnormality is supported by the presence of karyotypic aberrations in many cases; the occasional transformation of aplasias into myeloid neoplasms, typically myelodysplasia or acute myeloid leukemia; and the association with abnormally short telomeres. Some marrow insult (or a predisposition to DNA damage) presumably results in sufficient injury to limit the proliferative and differentiative capacity of stem cells. If the damage is extensive enough, aplastic anemia results. These two mechanisms are not mutually exclusive, since genetically altered stem cells might also express “neoantigens” that could serve as targets for a T-cell attack.

Morphology. The markedly hypocellular bone marrow is largely devoid of hematopoietic cells; often only fat cells, fibrous stroma, and scattered lymphocytes and plasma cells remain. Marrow aspirates often yield little material (a “dry tap”); hence, aplasia is best appreciated in marrow biopsies (Fig. 14-25). Other nonspecific pathologic changes are related to granulocytopenia and thrombocytopenia, such as mucocutaneous bacterial infections and abnormal bleeding, respectively. If the anemia necessitates multiple transfusions, systemic hemosiderosis can appear.

FIGURE 14-25 Aplastic anemia (bone marrow biopsy). Markedly hypocellular marrow contains mainly fat cells. A, Low power; B, high power.

(Courtesy of Dr. Steven Kroft, Department of Pathology, University of Texas Southwestern Medical School, Dallas, TX.)

Clinical Features.

Aplastic anemia can occur at any age and in either sex. The onset is usually insidious. Initial manifestations vary somewhat, depending on which cell line is predominantly affected, but pancytopenia ultimately appears, with the expected consequences. Anemia can cause progressive weakness, pallor, and dyspnea; thrombocytopenia is heralded by petechiae and ecchymoses; and neutropenia manifests as frequent and persistent minor infections or the sudden onset of chills, fever, and prostration. Splenomegaly is characteristically absent; if it is present, the diagnosis of aplastic anemia should be seriously questioned. The red cells are usually slightly macrocytic and normochromic. Reticulocytopenia is the rule.

The diagnosis rests on examination of a bone marrow biopsy. It is important to distinguish aplastic anemia from other causes of pancytopenia, such as “aleukemic” leukemia and myelodysplastic syndromes (Chapter 13), which can have identical clinical manifestations. In aplastic anemia, the marrow is hypocellular (and usually markedly so), whereas myeloid neoplasms are associated with hypercellular marrows filled with neoplastic progenitors.

The prognosis is variable.²⁰ Bone marrow transplantation is the treatment of choice in those with a suitable donor and provides a 5-year survival of over 75%. Older patients or those without suitable donors often respond well to immunosuppressive therapy.

Pathogenesis.

The autoantibodies, most often directed against platelet membrane glycoproteins IIb-IIIa or Ib-IX, can be demonstrated in the plasma and bound to the platelet surface in about 80% of patients. In the overwhelming majority of cases, the antiplatelet antibodies are of the IgG class.

As in autoimmune hemolytic anemias, antiplatelet antibodies act as opsonins that are recognized by IgG Fc receptors expressed on phagocytes (Chapter 6), leading to increased platelet destruction. The thrombocytopenia is usually markedly improved by splenectomy, indicating that the spleen is the major site of removal of opsonized platelets. The splenic red pulp is also rich in plasma cells, and part of the benefit of splenectomy (a common treatment for chronic ITP) may stem from the removal of a source of autoantibodies. In some instances the autoantibodies may also bind to and damage megakaryocytes, leading to decreases in platelet production that further exacerbate the thrombocytopenia.

Clinical Features.

Chronic ITP occurs most commonly in adult women typically under 40 years of age. The female-to-male ratio is 3 : 1. It is often insidious in onset and is characterized by bleeding into the skin and mucosal surfaces. Cutaneous bleeding is seen in the form of pinpoint hemorrhages (petechiae), which are especially prominent in the dependent areas where the capillary pressure is higher. Petechiae can become confluent, giving rise to ecchymoses. Often there is a history of easy bruising, nosebleeds, bleeding from the gums, and hemorrhages into soft tissues from relatively minor trauma. The disease may manifest first with melena, hematuria, or excessive menstrual flow. Subarachnoid hemorrhage and intracerebral hemorrhage are serious and sometimes fatal complications, but fortunately they are rare in treated patients. Splenomegaly and lymphadenopathy are uncommon in primary disease, and their presence should lead one to consider other diagnoses, such as ITP secondary to a B-cell neoplasm.

The clinical signs and symptoms are not specific but rather reflective of the thrombocytopenia. The findings of a low platelet count, normal or increased megakaryocytes in the bone marrow, and large platelets in the peripheral blood are taken as presumptive evidence of accelerated platelet destruction. The PT and PTT are normal. Tests for platelet autoantibodies are not widely available. Therefore, the diagnosis is one of exclusion and can be made only after other causes of thrombocytopenia, such as those listed inTable 14-9, have been ruled out.

Almost all patients respond to glucocorticoids (which inhibit phagocyte function), but many eventually relapse. In such individuals, splenectomy normalizes the platelet count in about two thirds of patients, but with the attendant in-creased risk of bacterial sepsis. Immunomodulatory agents such as intravenous immunoglobulin or anti-CD20 antibody (rituximab) are often effective in patients who relapse after splenectomy or for whom splenectomy is contraindicated.

Etiology and Pathogenesis.

At the outset, it must be emphasized that DIC is not a primary disease. It is a coagulopathy that occurs in the course of a variety of clinical conditions. In discussing the general mechanisms underlying DIC, it is useful to briefly review the normal process of blood coagulation and clot removal (see Chapter 4 for more details).

Clotting can be initiated by either of two pathways: (1) the extrinsic pathway, which is triggered by the release of tissue factor (“tissue thromboplastin”); and (2) the intrinsic pathway, which involves the activation of factor XII by surface contact with collagen or other negatively charged substances. Both pathways, through a series of intermediate steps, result in the generation of thrombin, which in turn converts fibrinogen to fibrin. At the site of injury, thrombin further augments local fibrin deposition by directly activating the intrinsic pathway and factors that inhibit fibrinolysis.

Once clotting is initiated, it is critically important that it be limited to the site of injury. Remarkably, as thrombin is swept away in the bloodstream and encounters normal vessels, it is converted to an anticoagulant through binding to thrombomodulin, a protein found on the surface of endothelial cells. The thrombin-thrombomodulin complex activates protein C, which is an important inhibitor of two procoagulants, factor V and factor VIII. Other activated coagulation factors are removed from the circulation by the liver, and as you will recall, the blood also contains several potent fibrinolytic factors, such as plasmin. These and additional checks and balances normally ensure that just enough clotting occurs at the right place and time.

From this brief review it should be clear that DIC could result from pathologic activation of the extrinsic and/or intrinsic pathways of coagulation or the impairment of clot-inhibiting mechanisms. Since the latter rarely constitute primary mechanisms of DIC, we will focus on the abnormal initiation of clotting.

Two major mechanisms trigger DIC: (1) release of tissue factor or thromboplastic substances into the circulation, and (2) widespread injury to the endothelial cells. Thromboplastic substances can be derived from a variety of sources, such as the placenta in obstetric complications and the cytoplasmic granules of acute promyelocytic leukemia cells (Chapter 13). Mucus released from certain adenocarcinomas can directly activate factor X, independent of factor VII.

Endothelial injury can initiate DIC in several ways. Injuries that cause endothelial cell necrosis expose the subendothelial matrix, leading to the activation of platelets and both arms of the coagulation pathway. However, even subtle endothelial injuries can unleash procoagulant activity. One mediator of such effects is TNF, which is implicated in DIC occurring with sepsis. TNF induces endothelial cells to express tissue factor on their cell surfaces and to decrease the expression of thrombomodulin, shifting the checks and balances that govern hemostasis towards coagulation. In addition, TNF upregulates the expression of adhesion molecules on endothelial cells, thereby promoting the adhesion of leukocytes, which can damage endothelial cells by releasing reactive oxygen species and preformed proteases. Widespread endothelial injury may also be produced by deposition of antigen-antibody complexes (e.g., systemic lupus erythematosus), temperature extremes (e.g., heat stroke, burns), or microorganisms (e.g., meningococci, rickettsiae). Even subtle endothelial injury can unleash procoagulant activity by enhancing membrane expression of tissue factor.

DIC is most likely to be associated with obstetric complications, malignant neoplasms, sepsis, and major trauma. The triggers in these conditions are often multiple and interrelated. For example, in bacterial infections endotoxins can injure endothelial cells and inhibit the expression of thrombomodulin directly or through production of TNF; stimulate the release of thromboplastins from inflammatory cells; and activate factor XII. Antigen-antibody complexes formed in response to the infection can activate the classical complement pathway, giving rise to complement fragments that secondarily activate both platelets and granulocytes. In massive trauma, extensive surgery, and severe burns, the major trigger is the release of tissue thromboplastins. In obstetric conditions, thromboplastins derived from the placenta, dead retained fetus, or amniotic fluid may enter the circulation. Hypoxia, acidosis, and shock, which often coexist in very ill patients, can also cause widespread endothelial injury, and supervening infections can complicate the problems further. Among cancers, acute promyelocytic leukemia and adenocarcinomas of the lung, pancreas, colon, and stomach are most frequently associated with DIC.

The possible consequences of DIC are twofold (Fig. 14-27). Firstly, there is widespread deposition of fibrin within the microcirculation. This leads to ischemia of the more severely affected or more vulnerable organs and a microangiopathic hemolytic anemia, which results from the fragmentation of red cells as they squeeze through the narrowed microvasculature. Secondly, the consumption of platelets and clotting factors and the activation of plasminogen leads to a hemorrhagic diathesis. Plasmin not only cleaves fibrin, but it also digests factors V and VIII, thereby reducing their concentration further. In addition, fibrin degradation products resulting from fibrinolysis inhibit platelet aggregation, fibrin polymerization, and thrombin. All of these derangements contribute to the hemostatic failure seen in DIC.

FIGURE 14-27 Pathophysiology of disseminated intravascular coagulation.

Clinical Features.

The onset can be fulminant, as in endotoxic shock or amniotic fluid embolism, or insidious and chronic, as in cases of carcinomatosis or retention of a dead fetus. Overall, about 50% of the affected are obstetric patients having complications of pregnancy. In this setting the disorder tends to be reversible with delivery of the fetus. About 33% of the affected patients have carcinomatosis. The remaining cases are associated with the various entities previously listed.

It is almost impossible to detail all the potential clinical presentations, but a few common patterns are worthy of description. These include microangiopathic hemolytic anemia; dyspnea, cyanosis, and respiratory failure; convulsions and coma; oliguria and acute renal failure; and sudden or progressive circulatory failure and shock. In general, acute DIC, associated with obstetric complications or major trauma, for example, is dominated by a bleeding diathesis, whereas chronic DIC, such as occurs in cancer patients, tends to present with thrombotic complications. The diagnosis is based on clinical observation and laboratory studies, including measurement of fibrinogen levels, platelets, the PT and PTT, and fibrin degradation products.

The prognosis is highly variable and largely depends on the underlying disorder. The only definitive treatment is to remove or treat the inciting cause. The management requires meticulous maneuvering between the Scylla of thrombosis and the Charybdis of bleeding diathesis. Administration of anticoagulants or procoagulants has been advocated in specific settings, but not without controversy.