Chapter 37: Hematopoietic Agents: Growth Factors, Minerals, and Vitamins

History. Modern concepts of hematopoietic cell growth and differentiation arose in the 1950s when cells from the spleen and marrow were shown to play an important role in the restoration of hematopoietic tissue in irradiated animals. In 1961, Till and McCulloch demonstrated that individual hematopoietic cells could form macroscopic hematopoietic colonies in the spleens of irradiated mice. Their work established the concept of discrete hematopoietic stem cells, which can be experimentally identified, albeit in retrospect (i.e., the presence of a multilineage clonal splenic colony appearing 11 days after transplantation implied that a single cell lodged and expanded into several cell lineages). This concept now has been expanded to include normal human marrow cells. Moreover, such cells now can be prospectively identified.

The basis for identifying soluble growth factors was provided by Sachs and independently by Metcalf, who developed clonal, in vitro assays for hematopoietic progenitor cells. Initially, such hematopoietic colonies developed only in the presence of conditioned culture medium from leukocytes or tumor cell lines. Individual growth factors then were isolated based on their activities in clonal in vitro assays. Many of these same assays were instrumental in purifying a hierarchy of progenitor cells committed to individual and combinations of mature blood cells (Akashi, 2000; Nakorn, 2003).

The existence of a circulating growth factor that controls red blood cell development was first suggested by the experiments of Paul Carnot in 1906. He observed an increase in the red-cell count in rabbits injected with serum obtained from anemic animals and postulated the existence of a factor that he called hemopoietin. However, it was not until the 1950s that Reissmann, Erslev, and Jacobsen and coworkers defined the origin and actions of the hormone, now called erythropoietin. Subsequently, extensive studies of erythropoietin were carried out in patients with anemia and polycythemia, leading to the purification of erythropoietin from urine and the subsequent cloning of the erythropoietin gene. The high-level expression of erythropoietin in cell lines has allowed for its purification and use in humans with anemia.

Similarly, the existence of specific leukocyte growth factors was suggested by the capacity of different conditioned culture media to induce the in vitro growth of colonies containing different combinations of granulocytes and monocytes. An activity that stimulated the production of both granulocytes and monocytes was purified from murine lung-conditioned medium, leading to cloning of granulocyte/macrophage colony-stimulating factor (G-CSF), first from mice (Gough et al., 1984) and subsequently from humans (Wong, 1985). Finding an activity that stimulated the exclusive production of neutrophils permitted the cloning of granulocyte colony-stimulating factor (G-CSF) (Welte et al., 1985). Subsequently, a megakaryocyte colony-stimulating factor termed thrombopoietin was purified and cloned (Kaushansky, 1998).

The growth factors that support lymphocyte growth were not identified using in vitro colony-forming assays but rather using assays that measured the capacity of the cytokine to promote lymphocyte proliferation in vitro. This permitted the identification of the growth-promoting properties of interleukin (Il)-7, Il-4, or Il-15 for all lymphocytes, B cells, or NK cells, respectively (Goodwin et al., 1989; Grabstein et al., 1994). Again, recombinant expression of these cDNAs permitted production of sufficient quantities of biologically active growth factors for clinical investigations, allowing for the demonstration of the potential clinical utility of such factors.

Recombinant human erythropoietin (epoetin alfa) is nearly identical to the endogenous hormone except for two subtle differences. First, the carbohydrate modification pattern of epoetin alfa differs slightly from the native protein, but this difference apparently does not alter kinetics, potency, or immunoreactivity of the drug. However, modern assays can detect these differences (Skibeli et al., 2001) and identify athletes who use the recombinant product for "blood doping." The second difference probably is related to the manufacturing process because one commercially available form of the drug was recently associated with the development of anti-recombinant erythropoietin antibodies that cross-react with the patient's own erythropoietin, potentially causing pure red-cell aplasia (Macdougall, 2004). Most of these cases were caused by one preparation of the drug shortly after albumin was removed from the formulation (Casadevall, 2003).

During initial therapy and after any dosage adjustment, the hematocrit is determined once a week (HIV-infected and cancer patients) or twice a week (renal failure patients) until it has stabilized in the target range and the maintenance dose has been established; the hematocrit then is monitored at regular intervals. If the hematocrit increases by >4 points in any 2-week period, the dose should be decreased. Due to the time required for erythropoiesis and the erythrocyte half-life, hematocrit changes lag behind dosage adjustments by 2-6 weeks. The dose of darbepoetin should be decreased if the hemoglobin increase exceeds 1 g/dL in any 2-week period because of the association of excessive rate of rise of hemoglobin with adverse cardiovascular events.

During hemodialysis, patients receiving epoetin alfa or darbepoetin may require increased anticoagulation. Serious thromboembolic events have been reported, including migratory thrombophlebitis, microvascular thrombosis, pulmonary embolism, and thrombosis of the retinal artery and temporal and renal veins. The risk of thrombotic events, including vascular access thromboses, was higher in adults with ischemic heart disease or congestive heart failure receiving epoetin alfa therapy with the goal of reaching a normal hematocrit (42%) than in those with a lower target hematocrit of 30%. A large meta-analysis indicated that there was a relative risk of 1.6 for developing venous thromboembolic events associated with erythropoietic therapies (Bennett et al., 2008). The higher risk of cardiovascular events from erythropoietic therapies may be associated with higher hemoglobin or higher rates of rise of hemoglobin. The hemoglobin level should be managed to avoid exceeding a target level of 12 g/dL. ESA use is associated with increased rates of cancer recurrence and decreased on-study survival in patients in whom the drugs are administered for cancer-induced or for chemotherapy-induced anemia. The magnitude of the effect was estimated as ∼17% based on a meta-analysis of nearly 14,000 cancer patients (Bohlius et al., 2008). The cause(s) of this effect is presently unclear, but some studies suggest that tumor cells bearing the erythropoietin receptor are more likely to be affected by the use of ESAs (Henke et al., 2003). Although epoetin alfa is not associated with direct pressor effects, blood pressure may rise, especially during the early phases of therapy when the hematocrit is increasing. ESAs should not be used in patients with preexisting uncontrolled hypertension. Patients may require initiation of, or increases in, antihypertensive therapy. Hypertensive encephalopathy and seizures have occurred in chronic renal failure patients treated with epoetin alfa. The incidence of seizures appears to be higher during the first 90 days of therapy with epoetin alfa in patients on dialysis (occurring in ∼2.5% of patients) when compared with subsequent 90-day periods. Headache, tachycardia, edema, shortness of breath, nausea, vomiting, diarrhea, injection site stinging, and flu-like symptoms (e.g., arthralgias and myalgias) also have been reported in conjunction with epoetin alfa therapy. Pure red-cell aplasia in association with neutralizing antibodies to native erythropoietin has been observed in patients treated with specific epoetin alfa formulations (see earlier discussion). Resistance to therapy can be caused by multiple factors (see following discussion).

Anemia of Chronic Renal Failure. Patients with anemia secondary to chronic kidney disease are ideal candidates for epoetin alfa therapy because the disorder represents a pure erythropoietin deficiency state. The response in predialysis, peritoneal dialysis, and hemodialysis patients depends on the severity of the renal failure, the erythropoietin dose and route of administration, and iron availability (Besarab et al., 1999, Kaufman et al., 1998). The subcutaneous route of administration is preferred over the intravenous route because absorption is slower and the amount of drug required is reduced by 20-40%.

The dose of epoetin alfa should be adjusted to obtain a gradual rise in the hematocrit over a 2- to 4-month period to a final hematocrit of 33-36%. Treatment to hematocrit levels >36% is not recommended because patients treated to a hematocrit >40% showed a higher incidence of myocardial infarction and death (Besarab et al., 1998). The drug should not be used to replace emergency transfusion in patients who need immediate correction of a life-threatening anemia.

Patients are started on doses of 80-120 units/kg of epoetin alfa, given subcutaneously, three times a week. It can be given on a once-weekly schedule, but somewhat more drug is required for an equivalent effect. If the response is poor, the dose should be progressively increased. The final maintenance dose of epoetin alfa can vary from as little as 10 units/kg to >300 units/kg, with an average dose of 75 units/kg, three times a week. Children <5 years of age generally require a higher dose. Resistance to therapy is common in patients who develop an inflammatory illness or become iron deficient, so close monitoring of general health and iron status is essential. Less common causes of resistance include occult blood loss, folic acid deficiency, carnitine deficiency, inadequate dialysis, aluminum toxicity, and osteitis fibrosa cystica secondary to hyperparathyroidism.

The most common side effect of epoetin alfa therapy is aggravation of hypertension, which occurs in 20-30% of patients and most often is associated with a rapid rise in hematocrit. Blood pressure usually can be controlled either by increasing antihypertensive therapy or ultrafiltration in dialysis patients or by reducing the epoetin alfa dose to slow the hematocrit response.

Darbepoetin alfa also is approved for use in patients who are anemic secondary to chronic kidney disease. The recommended starting dose is 0.45 μg/kg administered intravenously or subcutaneously once weekly, with dose adjustments depending on the response. Like epoetin alfa, side effects tend to occur when patients experience a rapid rise in hemoglobin concentration; a rise of <1 g/dL every 2 weeks generally has been considered safe.

Anemia in AIDS Patients. Epoetin alfa therapy has been approved for the treatment of HIV-infected patients, especially those on zidovudine therapy (Fischl et al., 1990). Excellent responses to doses of 100-300 units/kg, given subcutaneously three times a week, generally are seen in patients with zidovudine-induced anemia. In the face of advanced disease, marrow damage, and elevated serum erythropoietin levels (>500 IU/L), therapy is less effective.

Cancer-Related Anemias. Epoetin alfa therapy, 150 units/kg three times a week or 450-600 units/kg once a week, can reduce the transfusion requirement in cancer patients undergoing chemotherapy. Evidence-based guidelines for the therapeutic use of recombinant erythropoietin in patients with cancer have been published (Rizzo et al., 2002). Briefly, the guidelines recommend the use of epoetin alfa in patients with chemotherapy-associated anemia when hemoglobin levels fall below 10 g/dL, basing the decision to treat less severe anemia (Hb, 10-12 g/dL) on clinical circumstances. For anemia associated with hematologic malignancies, the guidelines support the use of recombinant erythropoietin in patients with low-grade myelodysplastic syndrome, although the evidence that the drug is effective in anemic patients with multiple myeloma, non-Hodgkin's lymphoma, or chronic lymphocytic leukemia not receiving chemotherapy is less robust. A baseline serum erythropoietin level may help to predict the response; most patients with blood levels >500 IU/L are unlikely to respond to any dose of the drug. Most patients treated with epoetin alfa experienced an improvement in their anemia, sense of well-being, and quality of life (Littlewood et al., 2001). This improved sense of well-being, particularly in cancer patients, may not be solely due to the rise in the hematocrit. Erythropoietin receptors have been demonstrated in cells of the central nervous system (CNS), and erythropoietin has been found to act as a cytoprotectant in several models of CNS ischemia. Thus high levels of the hormone may directly affect cancer patients' sense of well-being.

Darbepoetin alfa also has been tested in cancer patients undergoing chemotherapy and preliminary studies appear promising. However, recent case reports have suggested a direct effect of both epoetin alfa and darbepoetin alfa in stimulation of tumor cells. For example, patients with cancer of the head and neck randomized to receive recombinant erythropoietin had a statistically significant increase in likelihood of tumor progression during the duration of the study (Henke et al., 2003). A meta-analysis of a large number of patients and clinical trials estimates the risk at ∼10% higher than nontreated cancer patients (Bohlius et al., 2008). This finding is being evaluated by the FDA and warrants serious attention.

Surgery and Autologous Blood Donation. Epoetin alfa has been used perioperatively to treat anemia and reduce the need for erythrocyte transfusion. Patients undergoing elective orthopedic and cardiac procedures have been treated with 150-300 units/kg of epoetin alfa once daily for the 10 days preceding surgery, on the day of surgery, and for 4 days after surgery. As an alternative, 600 units/kg can be given on days 21, 14, and 7 before surgery, with an additional dose on the day of surgery. This can correct a moderately severe preoperative anemia (i.e., hematocrit 30-36%) and reduce the need for transfusion. Epoetin alfa also has been used to improve autologous blood donation (Goodnough et al., 1989). However, the potential benefit generally is small, and the expense is considerable. Patients treated for 3-4 weeks with epoetin alfa (300-600 units/kg twice a week) are able to donate only 1 or 2 more units than untreated patients, and most of the time this goes unused. Still, the ability to stimulate erythropoiesis for blood storage can be invaluable in the patient with multiple alloantibodies to red blood cells.

Other Uses. Epoetin alfa has received orphan drug status from the FDA for the treatment of the anemia of prematurity, HIV infection, and myelodysplasia. In the latter case, even very high doses >1000 units/kg two to three times a week have had limited success. The utility of very high-dose therapy in other hematologic disorders, such as sickle cell anemia, is still under study. Highly competitive athletes have used epoetin alfa to increase their hemoglobin levels ("blood doping") and improve performance. Unfortunately, this misuse of the drug has been implicated in the deaths of several athletes and is strongly discouraged.

The role of GM-CSF therapy in allogeneic transplantation is less clear. Its effect on neutrophil recovery is less pronounced in patients receiving prophylactic treatment for graft-versus-host disease (GVHD), and studies have failed to show a significant effect on transplant mortality, long-term survival, the appearance of GVHD, or disease relapse. However, it may improve survival in transplant patients who exhibit early graft failure (Nemunaitis et al., 1990). It also has been used to mobilize CD34-positive progenitor cells for peripheral blood stem cell collection for transplantation after myeloablative chemotherapy (Haas et al., 1990). Sargramostim has been used to shorten the period of neutropenia and reduce morbidity in patients receiving intensive cancer chemotherapy (Gerhartz et al., 1993). It also stimulates myelopoiesis in some patients with cyclic neutropenia, myelodysplasia, aplastic anemia, or AIDS-associated neutropenia.

Sargramostim (leukine) is administered by subcutaneous injection or slow intravenous infusion at doses of 125-500 μg/m² per day. Plasma levels of GM-CSF rise rapidly after subcutaneous injection and then decline with a t_1/2 of 2-3 hours. When given intravenously, infusions should be maintained over 3-6 hours. With the initiation of therapy, there is a transient decrease in the absolute leukocyte count secondary to margination and sequestration in the lungs. This is followed by a dose-dependent, biphasic increase in leukocyte counts over the next 7-10 days. Once the drug is discontinued, the leukocyte count returns to baseline within 2-10 days. When GM-CSF is given in lower doses, the response is primarily neutrophilic, whereas monocytosis and eosinophilia are observed at larger doses. After hematopoietic stem cell transplantation or intensive chemotherapy, sargramostim is given daily during the period of maximum neutropenia until a sustained rise in the granulocyte count is observed. Frequent blood counts are essential to avoid an excessive rise in the granulocyte count. The dose may be increased if the patient fails to respond after 7-14 days of therapy. However, higher doses are associated with more pronounced side effects, including bone pain, malaise, flu-like symptoms, fever, diarrhea, dyspnea, and rash. An acute reaction to the first dose, characterized by flushing, hypotension, nausea, vomiting, and dyspnea, with a fall in arterial oxygen saturation due to granulocyte sequestration in the pulmonary circulation occurs in sensitive patients. With prolonged administration, a few patients may develop a capillary leak syndrome, with peripheral edema and pleural and pericardial effusions. Other serious side effects include transient supraventricular arrhythmia, dyspnea, and elevation of serum creatinine, bilirubin, and hepatic enzymes.

Filgrastim is administered by subcutaneous injection or intravenous infusion over at least 30 minutes at doses of 1-20 μg/kg per day. The usual starting dose in a patient receiving myelosuppressive chemotherapy is 5 μg/kg per day. The distribution and clearance rate from plasma (t_1/2 of 3.5 hours) are similar for both routes of administration. As with GM-CSF therapy, filgrastim given daily after hematopoietic stem cell transplantation or intensive cancer chemotherapy will increase granulocyte production and shorten the period of severe neutropenia. Frequent blood counts should be obtained to determine the effectiveness of the treatment and guide dosage adjustment. In patients who received intensive myelosuppressive cancer chemotherapy, daily administration of G-CSF for ≥14-21 days may be necessary to correct the neutropenia. With less intensive chemotherapy, <7 days of treatment may suffice. In AIDS patients on zidovudine or patients with cyclic neutropenia, chronic G-CSF therapy often is required.

One indication for G-CSF presently under investigation is its use to increase the number of peripheral blood neutrophils in leukocyte donors. For many years it had been hoped that, like platelet transfusions for the bleeding associated with severe thrombocytopenia, neutrophil transfusion could diminish the infectious complications of neutropenia. However, given the short circulatory t_1/2 of neutrophils (∼6 hours) and the need for large numbers of cells, the practical collection of sufficient cell numbers has eluded hematologists. With few complications of therapy in >15 years of clinical experience, G-CSF now has been used to increase peripheral neutrophil counts in prospective donors and neutrophil transfusions (Hubel et al., 2002). Although initial results were modest, the therapy is likely to be optimized and greater efficacy is anticipated.

Adverse reactions to filgrastim include mild to moderate bone pain in patients receiving high doses over a protracted period, local skin reactions following subcutaneous injection, and rare cutaneous necrotizing vasculitis. Patients with a history of hypersensitivity to proteins produced by E. coli should not receive the drug. Marked granulocytosis, with counts >100,000/μL, can occur in patients receiving filgrastim over a prolonged period of time. However, this is not associated with any reported clinical morbidity or mortality and rapidly resolves once therapy is discontinued. Mild to moderate splenomegaly has been observed in patients on long-term therapy.

The therapeutic roles of other growth factors still need to be defined, although IL-3 and IL-6 have been removed from testing due to poor efficacy and/or significant toxicity. M-CSF may play a role in stimulating monocyte and macrophage production, although with significant side effects, including splenomegaly and thrombocytopenia. Stem cell factor (SCF) has been shown to augment peripheral blood mobilization of primitive hematopoietic progenitor cells (Moskowitz et al., 1997).

The drug is available in single-use vials containing 5 mg and is administered to patients at 25-50 μg/kg per day subcutaneously. Oprelvekin is approved for use in patients undergoing chemotherapy for nonmyeloid malignancies that displayed severe thrombocytopenia (platelet count <20,000/μL) on a prior cycle of the same chemotherapy, and it is administered until the platelet count returns to >100,000/μL. The major complications of therapy are fluid retention and associated cardiac symptoms, such as tachycardia, palpitation, edema, and shortness of breath; this is a significant concern in elderly patients and often requires concomitant therapy with diuretics. Fluid retention reverses upon drug discontinuation, but volume status should be carefully monitored in elderly patients, those with a history of heart failure, or those with preexisting fluid collections in the pleura, pericardium, or peritoneal cavity. Also reported are blurred vision, injection-site rash or erythema, and paresthesias.

In clinical trials, both drugs are safe in the patient populations selected for study. However, efficacy results using these agents have been mixed. In a small number of patients with gynecological cancers who were receiving carboplatin (Vadhan-Raj et al., 2000), rHuTPO therapy reduced the duration of severe thrombocytopenia and the need for platelet transfusions. In a similar study of patients treated with carboplatin plus cyclophosphamide, patients receiving a cycle of chemotherapy supplemented with G-CSF plus thrombopoietin had higher platelet counts at nadir and a shorter median duration of severe thrombocytopenia than they did after cycles of therapy supplemented only with G-CSF (Basser et al., 2000). When used to augment peripheral blood counts in preparation for platelet donation, a single dose of thrombopoietin in the platelet donors tripled their platelet counts, allowed for a threefold increase in the number of platelets that could be collected in a single apheresis, and led to a fourfold increase in the mean platelet count noted in transfusion recipients (Kuter et al., 2001). However, this particular regimen was associated with several instances of anti-recombinant thrombopoietin antibodies that cross-reacted with the native hormone, resulting in subsequent thrombocytopenia (Li et al., 2001).

In several studies, although the drug was safe, rHuMGDF was not effective. In two studies of patients treated for 7 days with standard aggressive therapy for acute leukemia, the addition of recombinant thrombopoietin failed to accelerate platelet recovery (Archimbaud et al., 1999). A similar lack of efficacy was seen when the drug was used following autologous peripheral blood stem cell transplantation (Bolwell et al., 2000). Failure to improve hematopoiesis in some of these trials may have resulted from the dosing regimen employed; the optimal dose and schedule of administration in various clinical settings need to be established. After a single bolus injection, platelet counts showed a detectable increase by day 4, peaked by 12-14 days, and then returned to normal over the next 4 weeks. The peak platelet response follows a log-linear dose response. Platelet activation and aggregation are not affected, and patients are not at increased risk of thromboembolic disease, unless the platelet count is allowed to rise to very high levels. These kinetics need to be taken into account when planning therapy in cancer chemotherapy patients.

Due to concerns over the immunogenicity of these agents, and to other considerations, efforts now are under way to develop small molecular mimics of recombinant thrombopoietin, discovered either through screening of phage display peptide libraries or of small organic molecules that have been developed for clinical use. Two of these agents are FDA approved for use in patients with immune thrombocytopenic purpura (ITP) who have failed to respond to more conventional treatments. Romiplostim (nplate) contains four copies of a small peptide that binds with high affinity to the thrombopoietin receptor, grafted onto an immunoglobulin scaffold. Romiplostim was found safe and efficacious in two randomized controlled studies in patients with ITP. Overall, ∼84% of patients responded to the drug with substantial increases in platelet levels, of which approximately half were durable (platelets >50,000/μL for 6 of the last 8 weeks of study) (Kuter et al., 2008). The drug is administered weekly by subcutaneous injection, starting with a dose of 1 μg/kg, titrated to a maximum of 10 μg/kg, until platelet count increases above 50,000/μL. Eltrombopag (promacta) is a small organic molecule that binds specifically to the thrombopoietin receptor and is administered orally. The safety and efficacy of eltrombopag were evaluated in two double-blind placebo-controlled clinical studies of >200 adult patients with chronic ITP who had completed at least one prior treatment course and who had severe thrombocytopenia (Bussel et al., 2007). These studies demonstrated that 70-81% of patients with ITP can be expected to respond to a 6-week course of 50-75 mg/day of eltrombopag. The recommended starting dose is 30 g per day, titrated to 75 mg depending on platelet response. Several of these agents are in clinical trials.

History. The modern understanding of iron metabolism began in 1937 with the work of McCance and Widdowson on iron absorption and excretion and Heilmeyer and Plotner's measurement of iron in plasma (Beutler, 2002). In 1947, Laurell described a plasma iron transport protein that he called transferrin (Laurell, 1951). Hahn and coworkers first used radioactive isotopes to measure iron absorption and define the role of the intestinal mucosa to regulate this function (Hahn, 1948). In the next decade, Huff and associates initiated isotopic studies of internal iron metabolism. The subsequent development of practical clinical measurements of serum iron, transferrin saturation, plasma ferritin, and red-cell protoporphyrin permitted the definition and detection of the body's iron store status and iron-deficient erythropoiesis. In 1994, Feder and colleagues identified the HFE gene, which is mutated in type 1 hemochromatosis, on the short arm of chromosome 6 at 6p21.3 (Feder et al., 1996). In 2000, Ganz and colleagues discovered a peptide produced by the liver, which was termed hepcidin (Park et al., 2001). Soon thereafter hepcidin was found to be the master regulator of iron homeostasis and to play a role in anemia of chronic disease (Ganz, 2003; Ganz and Nemeth, 2009).

Iron and the Environment. Iron exists in the environment largely as ferric oxide or hydroxide or as polymers. In this state, its biological availability is limited unless solubilized by acid or chelating agents. For example, bacteria and some plants produce high-affinity chelating agents that extract iron from the surrounding environment. Most mammals have little difficulty in acquiring iron; this is explained by an ample iron intake and perhaps also by a greater efficiency in absorbing iron. Humans, however, appear to be an exception. Although total dietary intake of elemental iron in humans usually exceeds requirements, the bioavailability of the iron in the diet is limited.

Cells regulate their expression of transferrin receptors and intracellular ferritin in response to the iron supply (De Domenico et al., 2008). The synthesis of apoferritin and transferrin receptors is regulated post-transcriptionally by two iron-regulatory proteins 1 and 2 (IRP1 and IRP2). Double knockout of the genes encoding these proteins is embryonic lethal, and conditional double knockout of these genes in the intestine results in cellular iron depletion and death of intestinal epithelial cells (Galy et al., 2008). These IRPs are cytosolic RNA-binding proteins that bind to iron regulating elements (IREs) present in the 5′ or 3′ untranslated regions of mRNA encoding apoferritin or the transferrin receptors, respectively. Binding of these IRPs to the 5′ IRE of apoferritin mRNA represses translation, whereas binding to the 3′ IRE of mRNA encoding the transferrin receptors enhances transcript stability, thereby increasing protein production. When iron is abundant, IRP2 undergoes rapid proteolysis and IRP1 is converted from a RNA-binding protein into aconitase, an enzyme that catalyzes the interconversion of citrate and isocitrate. This results in increased production of apoferritin and reduced production of transferrin receptors. Conversely, when iron is in short supply, these IRPs accumulate, thereby repressing translation of apoferritin while enhancing production of transferrin receptors.

In developed countries, the normal adult diet contains ∼6 mg of iron per 1000 calories, providing an average daily intake for adult men of between 12 and 20 mg and for adult women of between 8 and 15 mg. Foods high in iron (>5 mg/100 g) include organ meats such as liver and heart, brewer's yeast, wheat germ, egg yolks, oysters, and certain dried beans and fruits; foods low in iron (<1 mg/100 g) include milk and milk products and most nongreen vegetables. The content of iron in food is affected further by the manner of its preparation because iron may be added from cooking in iron pots.

Although the iron content of the diet obviously is important, of greater nutritional significance is the bioavailability of iron in food. Heme iron, which constitutes only 6% of dietary iron, is far more available and is absorbed independent of the diet composition; it therefore represents 30% of iron absorbed (Conrad and Umbreit, 2002).

The nonheme fraction nonetheless represents far the largest amount of dietary iron ingested by the economically underprivileged. In a vegetarian diet, nonheme iron is absorbed very poorly because of the inhibitory action of a variety of dietary components, particularly phosphates. Ascorbic acid and meat facilitate the absorption of nonheme iron. Ascorbate forms complexes with and/or reduces ferric to ferrous iron. Meat facilitates the absorption of iron by stimulating production of gastric acid; other effects also may be involved. Either of these substances can increase availability several fold. Thus assessment of available dietary iron should include both the amount of iron ingested and an estimate of its availability (Figure 37–4) (Monsen et al., 1978).

A comparison of iron requirements with available dietary iron is seen in Table 37–4. Obviously, pregnancy and infancy represent periods of negative balance. Menstruating women also are at risk, whereas iron balance in adult men and nonmenstruating women is reasonably secure. The difference between dietary supply and requirements is reflected in the size of iron stores, which are low or absent when iron balance is precarious and high when iron balance is favorable (Table 37–2). Thus in infants after the third month of life and in pregnant women after the first trimester, stores of iron are negligible. Menstruating women have approximately one-third the stored iron found in adult men, indicative of the extent to which the additional average daily loss of ∼0.5 mg of iron affects iron balance.

The recognition of iron deficiency rests on an appreciation of the sequence of events that lead to depletion of iron stores. A negative balance first results in a reduction of iron stores and eventually a parallel decrease in red-cell iron and iron-related enzymes (Figure 37–5). In adults, depletion of iron stores may be recognized by a plasma ferritin <12 μg/L and the absence of reticuloendothelial hemosiderin in the marrow aspirate. Iron-deficient erythropoiesis is identified by a decreased saturation of transferrin to <16% and/or by an increase above normal in red-cell protoporphyrin. Iron-deficiency anemia is associated with a recognizable decrease in the concentration of hemoglobin in blood. However, the physiological variation in hemoglobin levels is so great that only about half the individuals with iron-deficient erythropoiesis are identified from their anemia. Moreover, so-called normal hemoglobin and iron values in infancy and childhood are lower because of the more restricted supply of iron in young children (Dallman et al., 1980).

In mild iron deficiency, identifying the underlying cause is more important than any symptoms related to the deficiency state. Because of the frequency of iron deficiency in infants and in menstruating or pregnant women, the need for exhaustive evaluation of such individuals usually is determined by the severity of the anemia. However, iron deficiency in men or postmenopausal women necessitates a search for a site of bleeding.

Although the presence of microcytic anemia is the most common indicator of iron deficiency, laboratory tests—such as measurement of transferrin saturation, red-cell protoporphyrin, and plasma ferritin—are required to distinguish iron deficiency from other causes of microcytosis. Such measurements are particularly useful when circulating red cells are not yet microcytic because of the recent nature of blood loss, but iron supply nonetheless limits erythropoiesis. More difficult is the differentiation of true iron deficiency from iron-deficient erythropoiesis due to inflammation. In the latter condition, iron stores actually are increased, but the release of iron from reticuloendothelial cells is blocked, the concentration of iron in plasma is decreased, and the supply of iron to the erythroid marrow becomes inadequate. The increased stores of iron in this condition may be demonstrated directly by examination of an aspirate of marrow or may be inferred from determination of an elevated plasma concentration of ferritin.

Clinically, the effectiveness of iron therapy is best evaluated by tracking the reticulocyte response and the rise in the hemoglobin or the hematocrit. An increase in the reticulocyte count is not observed for at least 4-7 days after beginning therapy. A measurable increase in the hemoglobin level takes even longer. A decision as to the effectiveness of treatment should not be made for 3-4 weeks after the start of treatment. An increase of ≥20 g/L in the concentration of hemoglobin by that time should be considered a positive response, assuming that no other change in the patient's clinical status can account for the improvement and that the patient has not been transfused.

If the response to oral iron is inadequate, the diagnosis must be reconsidered. A full laboratory evaluation should be conducted, and poor compliance by the patient or the presence of a concurrent inflammatory disease must be explored. A source of continued bleeding obviously should be sought. If no other explanation can be found, an evaluation of the patient's ability to absorb oral iron should be considered. There is no justification for merely continuing oral iron therapy beyond 3-4 weeks if a favorable response has not occurred.

Once a response to oral iron is demonstrated, therapy should be continued until the hemoglobin returns to normal. Treatment may be extended if it is desirable to replenish iron stores. This may require a considerable period of time because the rate of absorption of iron by the intestine will decrease markedly as iron stores are reconstituted. The prophylactic use of oral iron should be reserved for patients at high risk, including pregnant women, women with excessive menstrual blood loss, and infants. Iron supplements also may be of value for rapidly growing infants who are consuming substandard diets and for adults with a recognized cause of chronic blood loss. Except for infants, in whom the use of supplemented formulas is routine, the use of over-the-counter mixtures of vitamins and minerals to prevent iron deficiency should be discouraged.

Other iron compounds have utility in fortification of foods. Reduced iron (metallic iron, elemental iron) is as effective as ferrous sulfate, provided that the material employed has a small particle size. Large-particle ferrum reductum and iron phosphate salts have a much lower bioavailability, and their use for the fortification of foods is undoubtedly responsible for some of the confusion concerning effectiveness. Ferric edetate has been shown to have good bioavailability and to have advantages for maintenance of the normal appearance and taste of food.

The amount of iron, rather than the mass of the total salt in iron tablets, is important. It also is essential that the coating of the tablet dissolve rapidly in the stomach. Surprisingly, because iron usually is absorbed in the upper small intestine, certain delayed-release preparations have been reported to be effective and have been said to be even more effective than ferrous sulfate when taken with meals. However, reports of absorption from such preparations vary. Because a number of forms of delayed-release preparations are on the market and information on their bioavailability is limited, the effectiveness of most such preparations must be considered questionable.

A variety of substances designed to enhance the absorption of iron has been marketed, including surface-acting agents, carbohydrates, inorganic salts, amino acids, and vitamins. When present in an amount of ≥200 mg, ascorbic acid increases the absorption of medicinal iron by at least 30%. However, the increased uptake is associated with a significant increase in the incidence of side effects; therefore, the addition of ascorbic acid seems to have little advantage over increasing the amount of iron administered. It is inadvisable to use preparations that contain other compounds with therapeutic actions of their own, such as vitamin B₁₂, folate, or cobalt, because the patient's response to the combination cannot easily be interpreted.

The average dose for the treatment of iron-deficiency anemia is ∼200 mg of iron per day (2-3 mg/kg), given in three equal doses of 65 mg. Children weighing 15-30 kg can take half the average adult dose; small children and infants can tolerate relatively large doses of iron, e.g., 5 mg/kg. The dose used is a compromise between the desired therapeutic action and the toxic effects. Prophylaxis and mild nutritional iron deficiency may be managed with modest doses. When the object is the prevention of iron deficiency in pregnant women, e.g., doses of 15 to 30 mg of iron per day are adequate to meet the 3- to 6-mg daily requirement of the last two trimesters. When the purpose is to treat iron-deficiency anemia, but the circumstances do not demand haste, a total dose of ∼100 mg (35 mg three times daily) may be used.

The responses expected for different dosage regimens of oral iron are given in Table 37–5. These effects are modified by the severity of the iron-deficiency anemia and by the time of ingestion of iron relative to meals. Bioavailability of iron ingested with food is probably one-half or one-third of that seen in the fasting subject (Grebe et al., 1975). Antacids also reduce iron absorption if given concurrently. It is always preferable to administer iron in the fasting state, even if the dose must be reduced because of GI side effects. For patients who require maximal therapy to encourage a rapid response or to counteract continued bleeding, as much as 120 mg of iron may be administered four times a day. Sustained high rates of red-cell production require an uninterrupted supply of iron, and oral doses should be spaced equally to maintain a continuous high concentration of iron in plasma.

The duration of treatment is governed by the rate of recovery of hemoglobin and the desire to create iron stores. The former depends on the severity of the anemia. With a daily rate of repair of 2 g of hemoglobin per liter of whole blood, the red-cell mass usually is reconstituted within 1-2 months. Thus an individual with a hemoglobin of 50 g per liter may achieve a normal complement of 150 g/L in ∼50 days, whereas an individual with a hemoglobin of 100 g/L may take only half that time. The creation of stores of iron requires many months of oral iron administration. The rate of absorption decreases rapidly after recovery from anemia, and after 3-4 months of treatment, stores may increase at a rate of not much more than 100 mg/month. Much of the strategy of continued therapy depends on the estimated future iron balance. Patients with an inadequate diet may require continued therapy with low doses of iron. If the bleeding has stopped, no further therapy is required after the hemoglobin has returned to normal. With continued bleeding, long-term, high-dose therapy clearly is indicated.

Iron Poisoning. Large amounts of ferrous salts are toxic, but fatalities are rare in adults. Most deaths occur in children, particularly between the ages of 12 and 24 months. As little as 1-2 g of iron may cause death, but 2-10 g usually is ingested in fatal cases. The frequency of iron poisoning relates to its availability in the household, particularly the supply that remains after a pregnancy. The colored sugar coating of many of the commercially available tablets gives them the appearance of candy. All iron preparations should be kept in childproof bottles.

Signs and symptoms of severe poisoning may occur within 30 minutes after ingestion or may be delayed for several hours. They include abdominal pain, diarrhea, or vomiting of brown or bloody stomach contents containing pills. Of particular concern are pallor or cyanosis, lassitude, drowsiness, hyperventilation due to acidosis, and cardiovascular collapse. If death does not occur within 6 hours, there may be a transient period of apparent recovery, followed by death in 12-24 hours. The corrosive injury to the stomach may result in pyloric stenosis or gastric scarring. Hemorrhagic gastroenteritis and hepatic damage are prominent findings at autopsy. In the evaluation of a child thought to have ingested iron, a color test for iron in the gastric contents and an emergency determination of the concentration of iron in plasma can be performed. If the latter is <63 μmol (3.5 mg/L), the child is not in immediate danger. However, vomiting should be induced when there is iron in the stomach, and an x-ray should be taken to evaluate the number of pills remaining in the small bowel (iron tablets are radiopaque). Iron in the upper GI tract can be precipitated by lavage with sodium bicarbonate or phosphate solution, although the clinical benefit is questionable. When the plasma concentration of iron is greater than the total iron-binding capacity (63 μmol; 3.5 mg/L), deferoxamine should be administered (Chapter 67). Shock, dehydration, and acid-base abnormalities should be treated in the conventional manner. Most important is the speed of diagnosis and therapy. With early effective treatment, the mortality from iron poisoning can be reduced from as high as 45% to ∼1%.

Four iron formulations are available in the U.S. These are iron dextran (dexferrum or infed), sodium ferric gluconate (ferrlecit), ferumoxytol (feraheme), and iron sucrose (venofer). Ferumoxytol is a semisynthetic carbohydrate-coated superparamagnetic iron oxide nanoparticle that is approved for treatment of iron-deficiency anemia in patients with chronic kidney disease (Balakrishnan et al., 2009). The indications for the iron dextran preparations include treatment of any patient with documented iron deficiency and intolerance or irresponsiveness to oral iron. In contrast, the indications for the ferric gluconate and iron sucrose are limited to patients with chronic kidney disease who have a documented iron deficiency, although broader applications are being advocated (Wish, 2008). Iron dextran is the only preparation that can be given as a total-dose infusion, which is the entire dose of elemental iron required to replace iron stores. However, ferric gluconate and iron sucrose are administered as a fixed dose every week for several weeks until total iron stores are replenished.

Iron Dextran. Iron dextran injection (infed or dexferrum) is a colloidal solution of ferric oxyhydroxide complexed with polymerized dextran (molecular weight, ∼180,000 Da) that contains 50 mg/mL of elemental iron. The use of low-molecular-weight iron dextran has reduced the incidence of toxicity relative to that observed with high molecular weight preparations (imferon) (Auerbach and Al Talib, 2008). Iron dextran can be administered by either intravenous (preferred) or intramuscular injection. When given by deep intramuscular injection, it is gradually mobilized via the lymphatics and transported to reticuloendothelial cells; the iron then is released from the dextran complex. Intravenous administration gives a more reliable response. Given intravenously in a dose <500 mg, the iron dextran complex is cleared exponentially with a plasma t_1/2 of 6 hours. When ≥1 g is administered intravenously as total dose therapy, reticuloendothelial cell clearance is constant at 10-20 mg/hour. This slow rate of clearance results in a brownish discoloration of the plasma for several days and an elevation of the serum iron for 1-2 weeks.

Once the iron is released from the dextran within the reticuloendothelial cells, it is either incorporated into stores or transported via transferrin to the erythroid marrow. Although a portion of the processed iron is rapidly made available to the marrow, a significant fraction is only gradually converted to usable iron stores. All of the iron eventually is released, although many months are required before the process is complete. During this time, the iron dextran stores in reticuloendothelial cells can confuse the clinician who attempts to evaluate the iron status of the patient.

Intramuscular injection of iron dextran should only be initiated after a test dose of 0.5 mL (25 mg of iron). If no adverse reactions are observed, the injections can proceed. The daily dose ordinarily should not exceed 0.5 mL (25 mg of iron) for infants weighing <4.5 kg (10 lb), 1 mL (50 mg of iron) for children weighing <9 kg (20 lb), and 2 mL (100 mg of iron) for other patients. Iron dextran should be injected only into the muscle mass of the upper outer quadrant of the buttock using a z-track technique (displacement of the skin laterally before injection). However, local reactions and the concern about malignant change at the site of injection (Weinbren et al., 1978) make intramuscular administration inappropriate except when the intravenous route is inaccessible.

A test injection of 0.5 mL of undiluted iron dextran or an equivalent amount (25 mg of iron) diluted in saline also should precede intravenous administration of a therapeutic dose of iron dextran. The patient should be observed for signs of immediate anaphylaxis and for an hour after injection for any signs of vascular instability or hypersensitivity, including respiratory distress, hypotension, tachycardia, or back or chest pain. When widely spaced total-dose infusion therapy is given, a test dose injection should be given before each infusion because hypersensitivity can appear at any time. Furthermore, the patient should be monitored closely throughout the infusion for signs of cardiovascular instability. Delayed hypersensitivity reactions also are observed, especially in patients with rheumatoid arthritis or a history of allergies. Fever, malaise, lymphadenopathy, arthralgias, and urticaria can develop days or weeks following injection and last for prolonged periods of time. Therefore, iron dextran should be used with extreme caution in patients with rheumatoid arthritis or other connective tissue diseases, and during the acute phase of an inflammatory illness. Once hypersensitivity is documented, iron dextran therapy must be abandoned.

Before initiating iron dextran therapy, the total dose of iron required to repair the patient's iron-deficient state should be calculated. Relevant factors are the hemoglobin deficit, the need to reconstitute iron stores, and continued excess losses of iron, as seen with hemodialysis and chronic GI bleeding. Iron dextran solution (50 mg/mL of elemental iron) can be administered undiluted in daily doses of 2 mL until the total dose is reached or given off label as a single total-dose infusion. In the latter case, the iron dextran should be diluted in 250-1000 mL of 0.9% saline and infused over an hour or more.

With repeated doses of iron dextran—especially multiple total-dose infusions such as those sometimes used in the treatment of chronic GI blood loss—accumulations of slowly metabolized iron dextran stores in reticuloendothelial cells can be impressive. The plasma ferritin level also can rise to levels associated with iron overload. Although disease-related hemochromatosis has been associated with an increased risk of infections and cardiovascular disease, this has not been shown to be true in hemodialysis patients treated with iron dextran (Owen, 1999). It seems prudent, however, to withhold the drug whenever the plasma ferritin rises above 800 μg/L.

Sodium Ferric Gluconate. Sodium ferric gluconate is an intravenous iron preparation with a molecular size of ∼295,000 Da and an osmolality of 990 mOsm/kg⁻¹ (Balakrishnan et al., 2009). Administration of ferric gluconate at doses ranging from 62.5-125 mg during hemodialysis is associated with transferrin saturation exceeding 100% (Zanen et al., 1996). Hemodialysis patients who had ferritin levels between 500 and 1200 ng/mL and transferrin saturations of ≤25% while undergoing treatment with erythropoietin had improved hemoglobin values following treatment with ferric gluconate, resulting in reduced requirements for erythropoietin (Kapoian et al., 2008).

Unlike iron dextran, which requires processing by macrophages that may require several weeks, ∼80% of sodium ferric gluconate is delivered to transferrin with in 24 hours. Sodium ferric gluconate also has a lower risk of inducing serious anaphylactic reactions than iron dextran (Sengolge et al., 2005). No deaths were reported with 25 million infusions of sodium ferric gluconate, whereas there were 31 infusion-related deaths reported from approximately half the number of patients treated with iron dextran (Faich and Strobos, 1999). Thus, sodium ferric gluconate has become the preferred agent for parenteral iron therapy. Currently iron dextran is reserved for noncompliant patients or for those who are seriously inconvenienced by the multiple infusions that may be required for treatment with sodium ferric gluconate or iron sucrose.

Iron Sucrose. Iron sucrose is complex of polynuclear iron (III)-hydroxide in sucrose with a molecular weight of 252,000 Da and an osmolality of 1316 mOsm/kg⁻¹ (Balakrishnan et al., 2009). Following intravenous injection, the complex is taken up by the reticuloendothelial system, where it dissociates into iron and sucrose. Iron sucrose is generally administered in daily amounts of 100-200 mg within a 14-day period to a total cumulative dose of 1000 mg. It can be administered repeatedly to hemodialysis patients as maintenance therapy without inducing hypersensitivity (Aronoff et al., 2004).

Like sodium ferric gluconate, iron sucrose appears to be better tolerated and to cause fewer adverse events than iron dextran (Hayat, 2008). This agent is FDA-approved for the treatment of iron deficiency in patients with chronic kidney disease. However, iron sucrose has been used effectively to treat iron deficiency observed in other clinical settings (al-Momen et al., 1996; Bodemar et al., 2004). However, one study reported that iron sucrose may be most likely of available parenteral iron preparations to induce renal tubular injury because of its high renal uptake (Zager et al., 2004), potentially resulting in tubulointerstitial damage with chronic repeated use (Agarwal, 2006).

Copper has redox properties similar to that of iron, which simultaneously is essential and potentially toxic to the cell (Kim et al., 2008). Cells have virtually no free copper. Instead copper is stored by metallothioneins and distributed by specialized chaperones to sites that make use of its redox properties (Lalioti et al., 2009). Transfer of copper to nascent cuproenzymes is performed by individual or collective activities of P-type ATPases, ATP7A and ATP7B, which are expressed in all tissues (Linz and Lutsenko, 2007). In mammals, the liver is the organ most responsible for the storage, distribution, and excretion of copper. Mutations in ATP7A or ATP7B that interfere with this function have been found responsible for Wilson's disease or Menkes syndrome (steely hair syndrome) (de Bie et al., 2007), respectively, which can result in life-threatening hepatic failure.

Copper deficiency in experimental animals interferes with the absorption of iron and its release from reticuloendothelial cells (Gambling et al., 2008). The associated microcytic anemia is related to a decrease in the availability of iron to the normoblasts, and perhaps even more importantly, to decreased mitochondrial production of heme. It may be that the specific defect in the latter case is a decrease in the activity of cytochrome oxidase. Other pathological effects involving the skeletal, cardiovascular, and nervous systems have been observed in copper-deficient experimental animals. In humans, the prominent findings have been leukopenia, particularly granulocytopenia, and anemia. Concentrations of iron in plasma are variable, and the anemia is not always microcytic. When a low plasma copper concentration is determined in the presence of leukopenia and anemia, a therapeutic trial with copper is appropriate. Daily doses up to 0.1 mg/kg of cupric sulfate have been given by mouth, or 1-2 mg per day may be added to the solution of nutrients for parenteral administration. The daily U.S. Recommended Daily Allowance (RDA) of copper ranges from 1,300 μg for nursing women to 200 μg for infants 0-6 months of age.

Oral therapy with pyridoxine is of proven benefit in correcting the sideroblastic anemias associated with the antituberculosis drugs isoniazid and pyrazinamide, which act as vitamin B₆ antagonists. A daily dose of 50 mg of pyridoxine completely corrects the defect without interfering with treatment, and routine supplementation of pyridoxine often is recommended (see Chapter 56). In contrast, if pyridoxine is given to counteract the sideroblastic abnormality associated with administration of levodopa, the effectiveness of levodopa in controlling Parkinson's disease is decreased. Pyridoxine therapy does not correct the sideroblastic abnormalities produced by chloramphenicol or lead.

Patients with idiopathic acquired sideroblastic anemia generally fail to respond to oral pyridoxine, and those individuals who appear to have a pyridoxine-responsive anemia require prolonged therapy with large doses of the vitamin, 50-500 mg per day. Unfortunately, the early enthusiasm for treatment with pyridoxine was not reinforced by results of later studies. Moreover, even when a patient with sideroblastic anemia responds, the improvement is only partial because both the ringed sideroblasts and the red-cell defect persist, and the hematocrit rarely returns to normal. Nonetheless, in view of the low toxicity of oral pyridoxine, a therapeutic trial with pyridoxine is appropriate.

As shown in studies of normal subjects, oral pyridoxine in a dosage of 100 mg three times daily produces a maximal increase in red-cell pyridoxine kinase and the major pyridoxal phosphate–dependent enzyme glutamic-aspartic aminotransferase (Solomon and Hillman, 1978). For an adequate therapeutic trial, the drug is administered for at least 3 months while the response is monitored by measuring the reticulocyte index and the hemoglobin concentration. The occasional patient who is refractory to oral pyridoxine may respond to parenteral administration of pyridoxal phosphate. However, oral pyridoxine in doses of 200-300 mg per day produces intracellular concentrations of pyridoxal phosphate equal to or greater than those generated by therapy with the phosphorylated vitamin.

History. The discovery of vitamin B₁₂ and folic acid is a dramatic story that started >180 years ago and includes two discoveries that won the Nobel Prize. The first descriptions of what must have been megaloblastic anemias came from the work of Combe and Addison, who published a series of case reports beginning in 1824. It still is common practice to describe megaloblastic anemia as Addisonian pernicious anemia. Although Combe suggested that the disorder might have some relationship to digestion, it was Austin Flint who in 1860 first described the severe gastric atrophy and called attention to its possible relationship to the anemia.

After the observation by Whipple in 1925 that liver is a source of a potent hematopoietic substance for iron-deficient dogs, Minot and Murphy carried out Nobel Prize–winning experiments that demonstrated the effectiveness of the feeding of liver to reverse pernicious anemia. Soon thereafter, Castle defined the need for both intrinsic factor, a substance secreted by the parietal cells of the gastric mucosa, and extrinsic factor, the vitamin-like material provided by crude liver extracts. Nearly 20 years passed before Rickes and coworkers and Smith and Parker isolated and crystallized vitamin B₁₂; Dorothy Hodgkin received the Nobel Prize for determining its x-ray crystal structure.

As attempts were being made to purify extrinsic factor, Wills and her associates described a macrocytic anemia in women in India that responded to a factor present in crude liver extracts but not in the purified fractions known to be effective in pernicious anemia. This factor, first called Wills' factor and later vitamin M, is now known to be folic acid. The term folic acid was coined by Mitchell and coworkers in 1941, after its isolation from leafy vegetables.

More recent work has shown that neither vitamin B₁₂ nor folic acid as purified from foodstuffs is the active coenzyme in humans. During extraction, active labile forms are converted to stable congeners of vitamin B₁₂ and folic acid, cyanocobalamin and pteroylglutamic acid, respectively. These congeners must then be modified in vivo to be effective. Although a great deal has been learned about the intracellular metabolic pathways in which these vitamins function as required cofactors, many questions remain, among them: what is the relationship of vitamin B₁₂ deficiency to the neurological abnormalities that occur with this disorder?

Chemistry. The structural formula of vitamin B₁₂ is shown in Figure 37–7. These are the three major portions of the molecule:

1. A planar group or corrin nucleus—a porphyrin-like ring structure with four reduced pyrrole rings (A-D in Figure 37–7) linked to a central cobalt atom and extensively substituted with methyl, acetamide, and propionamide residues.

2. A 5,6-dimethylbenzimidazolyl nucleotide, which links almost at right angles to the corrin nucleus with bonds to the cobalt atom and to the propionate side chain of the C pyrrole ring.

3. A variable R group—the most important of which are found in the stable compounds cyanocobalamin and hydroxocobalamin and the active coenzymes methylcobalamin and 5-deoxyadenosylcobalamin.

The terms vitamin B₁₂ and cyanocobalamin are used interchangeably as generic terms for all of the cobamides active in humans. Preparations of vitamin B₁₂ for therapeutic use contain either cyanocobalamin or hydroxocobalamin because only these derivatives remain active after storage.

The supply of vitamin B₁₂ available for tissues is directly related to the size of the hepatic storage pool and the amount of vitamin B₁₂ bound to transcobalamin II (Figure 37–8). The plasma concentration of vitamin B₁₂ is the best routine measure of B₁₂ deficiency, and normally ranges from 150-660 pmol (∼200-900 pg/mL). Deficiency should be suspected whenever the concentration falls below 150 pmol. The correlation is excellent except when the plasma concentrations of transcobalamin I and III are increased, as occurs with hepatic disease or a myeloproliferative disorder. Inasmuch as the vitamin B₁₂ bound to these transport proteins is relatively unavailable to cells, tissues can become deficient when the concentration of vitamin B₁₂ in plasma is normal or even high. In subjects with congenital absence of transcobalamin II, megaloblastic anemia occurs despite relatively normal plasma concentrations of vitamin B₁₂; the anemia will respond to parenteral doses of vitamin B₁₂ that exceed the renal clearance (Hakami et al., 1971).

Defects in intracellular metabolism of vitamin B₁₂ have been reported in children with methylmalonic aciduria and homocystinuria. Potential mechanisms include an incapacity of cells to transport vitamin B₁₂ or accumulate the vitamin because of a failure to synthesize an intracellular acceptor, a defect in the formation of deoxyadenosylcobalamin, or a congenital lack of methylmalonyl CoA isomerase.

Cyanocobalamin injection is safe when given by the intramuscular or deep subcutaneous route, but it should never be given intravenously. There have been rare reports of transitory exanthema and anaphylaxis after injection. If a patient reports a previous sensitivity to injections of vitamin B₁₂, an intradermal skin test should be performed before the full dose is administered.

Cyanocobalamin is administered in doses of 1-1000 μg. Tissue uptake, storage, and utilization depend on the availability of transcobalamin II (discussed earlier). Doses >100 μg are cleared rapidly from plasma into the urine, and administration of larger amounts of vitamin B₁₂ will not result in greater retention of the vitamin. Administration of 1000 μg is of value in the performance of the Schilling test. After isotopically labeled vitamin B₁₂ is administered orally, the compound that is absorbed can be quantitatively recovered in the urine if 1000 μg of cyanocobalamin is administered intramuscularly. This unlabeled material saturates the transport system and tissue binding sites, so >90% of the labeled and unlabeled vitamin is excreted during the next 24 hours.

A number of multivitamin preparations are marketed either as nutritional supplements or for the treatment of anemia. Many of these contain up to 80 μg of cyanocobalamin without or with intrinsic factor concentrate prepared from the stomachs of hogs or other domestic animals. One oral unit of intrinsic factor is defined as that amount of material that will bind and transport 15 μg of cyanocobalamin. Most multivitamin preparations supplemented with intrinsic factor contain 0.5 oral unit per tablet. Although the combination of oral vitamin B₁₂ and intrinsic factor would appear to be ideal for patients with an intrinsic factor deficiency, such preparations are not reliable. With prolonged therapy, some patients become refractory to oral intrinsic factor, perhaps related to production of an intraluminal antibody against the hog protein. Patients taking such preparations must be reevaluated at periodic intervals for recurrence of pernicious anemia.

Hydroxocobalamin given in doses of 100 μg intramuscularly has been reported to have a more sustained effect than cyanocobalamin, with a single dose maintaining plasma vitamin B₁₂ concentrations in the normal range for up to 3 months. However, some patients show reductions of the concentration of vitamin B₁₂ in plasma within 30 days, similar to that seen after cyanocobalamin. Furthermore, the administration of hydroxocobalamin has resulted in the formation of antibodies to the transcobalamin II–vitamin B₁₂ complex.

Treatment of the Acutely Ill Patient. The therapeutic approach depends on the severity of the patient's illness. In uncomplicated pernicious anemia, in which the abnormality is restricted to a mild or moderate anemia without leukopenia, thrombocytopenia, or neurological signs or symptoms, the administration of vitamin B₁₂ alone will suffice. Moreover, therapy may be delayed until other causes of megaloblastic anemia have been excluded and sufficient studies of GI function have been performed to reveal the underlying cause of the disease. In this situation, a therapeutic trial with small amounts of parenteral vitamin B₁₂ (1-10 μg per day) can confirm the presence of an uncomplicated vitamin B₁₂ deficiency.

In contrast, patients with neurological changes or severe leukopenia or thrombocytopenia associated with infection or bleeding require emergency treatment. The older individual with a severe anemia (hematocrit <20%) is likely to have tissue hypoxia, cerebrovascular insufficiency, and congestive heart failure. Effective therapy must not wait for detailed diagnostic tests. Once the megaloblastic erythropoiesis has been confirmed and sufficient blood collected for later measurements of vitamin B₁₂ and folic acid, the patient should receive intramuscular injections of 100 μg of cyanocobalamin and 1-5 mg of folic acid. For the next 1-2 weeks the patient should receive daily intramuscular injections of 100 μg of cyanocobalamin, together with a daily oral supplement of 1 to 2 mg of folic acid. Because an effective increase in red-cell mass will not occur for 10-20 days, the patient with a markedly depressed hematocrit and tissue hypoxia also should receive a transfusion of 2-3 units of packed red blood cells. If congestive heart failure is present, diuretics can be administered to prevent volume overload.

Patients usually report an increased sense of well-being within the first 24 hours of the initiation of therapy. Objectively, memory and orientation can improve dramatically, although full recovery of mental function may take months, or it may never occur. In addition, even before an obvious hematologic response is apparent, the patient may report an increase in strength, a better appetite, and reduced soreness of the mouth and tongue.

The first objective hematologic change is the disappearance of the megaloblastic morphology of the bone marrow. As the ineffective erythropoiesis is corrected, the concentration of iron in plasma falls dramatically as the metal is used in the formation of hemoglobin. This usually occurs within the first 48 hours. Full correction of precursor maturation in marrow with production of an increased number of reticulocytes begins about the second or third day and peaks 3-5 days later. When the anemia is moderate to severe, the maximal reticulocyte index will be between three and five times the normal value (i.e., a reticulocyte count of 20-40%). The ability of the marrow to sustain a high rate of production determines the rate of recovery of the hematocrit. Patients with complicating iron deficiency, an infection or other inflammatory state, or renal disease may be unable to correct their anemia. Therefore it is important to monitor the reticulocyte index over the first several weeks. If it does not continue at elevated levels while the hematocrit is <35%, plasma concentrations of iron and folic acid should again be determined and the patient reevaluated for an illness that could inhibit the response of the marrow.

The degree and rate of improvement of neurological signs and symptoms depend on the severity and the duration of the abnormalities. Those that have been present for only a few months usually disappear relatively rapidly. When a defect has been present for many months or years, full return to normal function may never occur.

Long-Term Therapy with Vitamin B₁₂. Once begun, vitamin B₁₂ therapy must be maintained for life. This fact must be impressed on the patient and family, and a system must be established to guarantee continued monthly injections of cyanocobalamin.

Intramuscular injection of 100 μg of cyanocobalamin every 4 weeks is sufficient to maintain a normal concentration of vitamin B₁₂ in plasma and an adequate supply for tissues. Patients with severe neurological symptoms and signs may be treated with larger doses of vitamin B₁₂ in the period immediately after the diagnosis. Doses of 100 μg per day or several times per week may be given for several months with the hope of encouraging faster and more complete recovery. It is important to monitor vitamin B₁₂ concentrations in plasma and to obtain peripheral blood counts at intervals of 3-6 months to confirm the adequacy of therapy. Because refractoriness to therapy can develop at any time, evaluation must continue throughout the patient's life. Intranasal preparations are available for maintenance following normalization of vitamin B₁₂–deficient patients without nervous system involvement (calomist, nascobal) or as a supplement for vitamin B₁₂ deficiencies of various etiologies (nascobal).

Other Therapeutic Uses of Vitamin B₁₂. Vitamin B₁₂ has been used in the therapy of a number of conditions, including trigeminal neuralgia, multiple sclerosis and other neuropathies, various psychiatric disorders, poor growth or nutrition, and as a "tonic" for patients complaining of tiredness or easy fatigue. There is no evidence for the validity of such therapy in any of these conditions. Maintenance therapy with vitamin B₁₂ has been used with some apparent success in the treatment of children with methylmalonic aciduria.

Chemistry and Metabolic Functions. The structural formula of pteroylglutamic acid (PteGlu) is shown in Figure 37–9. Major portions of the molecule include a pteridine ring linked by a methylene bridge to para-aminobenzoic acid, which is joined by an amide linkage to glutamic acid. Although pteroylglutamic acid is the common pharmaceutical form of folic acid, it is neither the principal folate congener in food nor the active coenzyme for intracellular metabolism. After absorption, PteGlu is rapidly reduced at the 5, 6, 7, and 8 positions to tetrahydrofolic acid (H₄PteGlu), which then acts as an acceptor of a number of one-carbon units. These are attached at either the 5 or the 10 position of the pteridine ring or may bridge these atoms to form a new five-membered ring. The most important forms of the coenzyme that are synthesized by these reactions are listed in Figure 37–9. Each plays a specific role in intracellular metabolism, summarized as follows (see "Relationships Between Vitamin B12 and Folic Acid," as well as Figure 37–6).

Conversion of Homocysteine to Methionine. This reaction requires CH₃H₄PteGlu as a methyl donor and uses vitamin B₁₂ as a cofactor.

Conversion of Serine to Glycine. This reaction requires tetrahydrofolate as an acceptor of a methylene group from serine and uses pyridoxal phosphate as a cofactor. It results in the formation of 5,10-CH₂H₄PteGlu, an essential coenzyme for the synthesis of thymidylate.

Synthesis Of thymidylate. 5,10-CH₂H₄PteGlu donates a methylene group and reducing equivalents to deoxyuridylate for the synthesis of thymidylate—a rate-limiting step in DNA synthesis.

Histidine Metabolism. H₄PteGlu also acts as an acceptor of a formimino group in the conversion of formiminoglutamic acid to glutamic acid.

Synthesis of Purines. Two steps in the synthesis of purine nucleotides require the participation of 10-CHOH₄PteGlu as a formyl donor in reactions catalyzed by ribotide transformylases: the formylation of glycinamide ribonucleotide and the formylation of 5-aminoimidazole-4-carboxamide ribonucleotide. By these reactions, carbon atoms at positions 8 and 2, respectively, are incorporated into the growing purine ring.

Utilization or generation of formate. This reversible reaction uses H₄PteGlu and 10-CHOH₄PteGlu.

Generally, a standard U.S. diet provides 50-500 μg of absorbable folate per day, although individuals with high intakes of fresh vegetables and meats will ingest as much as 2 mg per day. In the normal adult, the recommended daily intake is 400 μg; pregnant or lactating women and patients with high rates of cell turnover (such as patients with a hemolytic anemia) may require 500-600 μg or more per day. For the prevention of neural tube defects, a daily intake of at least 400 μg of folate in food or in supplements beginning a month before pregnancy and continued for at least the first trimester is recommended. Folate supplementation also is being considered in patients with elevated levels of plasma homocysteine.

Once absorbed, folate is transported rapidly to tissues as CH₃H₄PteGlu. Although certain plasma proteins do bind folate derivatives, they have a greater affinity for nonmethylated analogs. The role of such binding proteins in folate homeostasis is not well understood. An increase in binding capacity is detectable in folate deficiency and in certain disease states, such as uremia, cancer, and alcoholism, but how binding affects transport and tissue supply requires further investigation.

A constant supply of CH₃H₄PteGlu is maintained by food and by an enterohepatic cycle of the vitamin. The liver actively reduces and methylates PteGlu (and H₂ or H₄PteGlu) and then transports the CH₃H₄PteGlu into bile for reabsorption by the gut and subsequent delivery to tissues. This pathway may provide ≥200 μg of folate each day for recirculation to tissues. The importance of the enterohepatic cycle is suggested by animal studies that show a rapid reduction of the plasma folate concentration after either drainage of bile or ingestion of alcohol, which apparently blocks the release of CH₃H₄PteGlu from hepatic parenchymal cells.

Folate deficiency has been implicated in the incidence of neural tube defects, including spina bifida, encephaloceles, and anencephaly. This is true even in the absence of folate-deficient anemia or alcoholism. A less-than-adequate intake of folate also can result in elevations in plasma homocysteine (Green and Miller, 1999). Because even moderate hyperhomocysteinemia is considered an independent risk factor for coronary artery and peripheral vascular disease and for venous thrombosis, the role of folate as a methyl donor in the homocysteine-to-methionine conversion is getting increased attention. Patients who are heterozygous for one or another enzymatic defect and have high normal to moderate elevations of plasma homocysteine may improve with folic acid therapy.

Folate deficiency is recognized by its impact on the hematopoietic system. As with vitamin B₁₂, this fact reflects the increased requirement associated with high rates of cell turnover. The megaloblastic anemia that results from folate deficiency cannot be distinguished from that caused by vitamin B₁₂ deficiency. This finding is to be expected because of the final common pathway of the major intracellular metabolic roles of the two vitamins. At the same time, folate deficiency is rarely if ever associated with neurological abnormalities. Thus the observation of characteristic abnormalities in vibratory and position sense and in motor and sensory pathways is incompatible with an isolated deficiency of folic acid.

After deprivation of folate, megaloblastic anemia develops much more rapidly than it does following interruption of vitamin B₁₂ absorption (e.g., gastric surgery). This observation reflects the fact that body stores of folate are limited. Although the rate of induction of megaloblastic erythropoiesis may vary, a folate-deficiency state may appear in 1-4 weeks, depending on the individual's dietary habits and stores of the vitamin.

Folate deficiency is best diagnosed from measurements of folate in plasma and in red cells. However, an empirical trial of folate in cases of suspected deficiency has been proposed as more cost effective (Robinson and Mladenovic, 2001). Indeed, the concentration of folate in plasma is extremely sensitive to changes in dietary intake of the vitamin and the influence of inhibitors of folate metabolism or transport, such as alcohol. Normal folate concentrations in plasma range from 9-45 nmol (4-20 ng/mL); <9 nmol is considered folate deficient. The plasma folate concentration rapidly falls to values indicative of deficiency within 24-48 hours of steady ingestion of alcohol (Eichner and Hillman 1973). The plasma folate concentration will revert quickly to normal once such ingestion is stopped, even while the marrow is still megaloblastic. Such rapid fluctuations detract from the clinical usefulness of the plasma folate concentration. The amount of folate in red cells or the adequacy of stores in lymphocytes (as measured by the deoxyuridine suppression test) may be used to diagnose a longstanding deficiency of folic acid. A positive result on either test shows that the deficiency must have existed for a sufficient time to allow the production of a population of cells with deficient folate stores.

Folic acid is marketed as oral tablets containing pteroylglutamic acid or L-methylfolate, as an aqueous solution for injection (5 mg/mL), and in combination with other vitamins and minerals.

Folinic acid (leucovorin calcium, citrovorum factor) is the 5-formyl derivative of tetrahydrofolic acid. The principal therapeutic uses of folinic acid are to circumvent the inhibition of dihydrofolate reductase as a part of high-dose methotrexate therapy and to potentiate fluorouracil in the treatment of colorectal cancer (Chapter 61). It also has been used as an antidote to counteract the toxicity of folate antagonists such as pyrimethamine or trimethoprim. Although it can be used to treat any folate-deficient state, folinic acid provides no advantage over folic acid, is more expensive, and therefore is not recommended. A single exception is the megaloblastic anemia associated with congenital dihydrofolate reductase deficiency. Leucovorin should never be used for the treatment of pernicious anemia or other megaloblastic anemias secondary to a deficiency of vitamin B₁₂. Just as is seen with folic acid, its use can result in an apparent response of the hematopoietic system, but neurological damage may occur or progress if already present. Leucovorin is marketed as oral tablets and as products for intravenous infusion.

Untoward Effects. There have been rare reports of reactions to parenteral injections of folic acid and leucovorin. If a patient describes a history of a reaction before the drug is given, caution should be exercised. Oral folic acid usually is not toxic. Even with doses as high as 15 mg/day, there have been no substantiated reports of side effects. Folic acid in large amounts may counteract the antiepileptic effect of phenobarbital, phenytoin, and primidone, and increase the frequency of seizures in susceptible children (Reynolds, 1968). Although some studies have not supported these contentions, the FDA recommends that oral tablets of folic acid be limited to strengths of ≤1 mg.

Treatment of the Acutely Ill Patient. As described in detail in the section on vitamin B₁₂, treatment of the patient who is acutely ill with megaloblastic anemia should begin with intramuscular injections of vitamin B₁₂ and folic acid. Inasmuch as the patient requires therapy before the exact cause of the disease has been defined, it is important to avoid the potential problem of a combined deficiency of vitamin B₁₂ and folic acid. When the patient is deficient in both, therapy with only one vitamin will not provide an optimal response. Longstanding nontropical sprue is one example of a disease in which combined deficiency of B₁₂ and folate is common. When indicated, vitamin B₁₂ (100 μg) and folic acid (1-5 mg) should be administered intramuscularly, and the patient should then be maintained on daily oral supplements of 1-2 mg of folic acid for the next 1-2 weeks.

Oral administration of folate generally is satisfactory for patients who are not acutely ill, regardless of the cause of the deficiency state. Even the patient with tropical or nontropical sprue and a demonstrable defect in absorption of folic acid will respond adequately to such therapy. Abnormalities in the activity of pteroyl-γ-glutamyl carboxypeptidase and the function of mucosal cells will not prevent passive diffusion of sufficient amounts of PteGlu across the mucosal barrier if the dosage is adequate, and continued ingestion of alcohol or other drugs also will not prevent an adequate therapeutic response. The effects of most inhibitors of folate transport or dihydrofolate reductase are overcome easily by administration of pharmacological doses of the vitamin. Folinic acid is the appropriate form of the vitamin for use in chemotherapeutic protocols, including "rescue" from methotrexate (Chapter 61). Perhaps the only situation in which oral administration of folate will be ineffective is when vitamin C is severely deficient. Patients with scurvy may suffer from a megaloblastic anemia despite increased intake of folate and normal or high concentrations of the vitamin in plasma and cells.

The therapeutic response may be monitored by study of the hematopoietic system in a fashion identical to that described for vitamin B₁₂. Within 48 hours of the initiation of appropriate therapy, megaloblastic erythropoiesis disappears, and as efficient erythropoiesis begins, the concentration of iron in plasma falls to normal or below-normal values. The reticulocyte count begins to rise on the second or third day and reaches a peak by the fifth to seventh days; the reticulocyte index reflects the proliferative state of the marrow. Finally, the hematocrit begins to rise during the second week.

It is possible to use the pattern of recovery as the basis for a therapeutic trial. For this purpose, the patient should receive a daily parenteral injection of 50-100 μg of folic acid. Administration of doses >100 μg/day entails the risk of inducing a hematopoietic response in patients who are deficient in vitamin B₁₂, and oral administration of the vitamin may be unreliable because of intestinal malabsorption. A number of other complications also may interfere with the therapeutic trial. The patient with sprue and deficiencies of other vitamins or iron may fail to respond because of these inadequacies. In cases of alcoholism, the presence of hepatic disease, inflammation, or iron deficiency can blunt the proliferative response of the marrow and prevent the correction of the anemia. For these reasons, the therapeutic trial for the evaluation of the patient with a potential deficiency of folic acid has not gained great popularity.