Use of Recombinant Human Erythropoietin as an Antianemic and Performance Enhancing Drug

W. Jelkmann*

 

Institut für Physiologie, Medizinische Universität zu Lübeck, Ratzeburger Allee 160, 23538 Lübeck, Germany

*Address correspondence to this author at the Institut für Physiologie, Medizinische Universität zu Lübeck, Ratzeburger Allee 160, 23538 Lübeck, Germany; Tel: +49-451-500-4150; Fax: +49-451-500-4151; E-mail: Jelkmann@physio.mu-Luebeck.de

Abstract: The glycoprotein hormone erythropoietin is an essential viability and growth factor for the erythrocytic progenitors in the bone marrow. Tissue hypoxia is the main stimulus for the synthesis of the hormone in the kidneys and the liver. Endogenous erythropoietin and recombinant human erythropoietin (rHu-EPO) are similar with respect to their biological and chemical properties except for some microheterogeneities in their 4 carbohydrate chains. Generic products and alternatives to rHu-EPO are in development. Renal anemia can be corrected by rHu-EPO in a dose-dependent and predictable way without major side effects apart from a possible increase in arterial blood pressure. The optimal target hematocrit still needs to be defined. There are rare reports of antibody formation towards rHu-EPO in humans. Patients suffering from non-renal anemias may also benefit from the prescription of rHu-EPO. The drug has been approved for treatment of tumor patients with platinum-induced anemia. The cost-effectiveness and medical justification of the administration of rHu-EPO in tumor patients with respect to its positive effects on tumor oxygenation, tumor growth inhibition and support of chemo- and radiotherapy is still a matter of debate. In surgical patients, the pharmacological application of rHu-EPO can increase the yield of blood units in autologous blood donation programs and lower the severity and duration of postoperative anemia, if applicated some days prior to surgery. While rHu-EPO is a godsend in medical practice, its abuse as an performance enhancing drug by athletes in endurance sports is an unethical and potentially dangerous procedure. Unequivocal methods for detection of rHu-EPO doping still need to be established.

Introduction

     Hematocrit and the concentration of hemoglobin in blood are normally maintained constant. About l % of the red cell mass is renewed each day. Anemic persons suffer from tissue hypoxia. They present with fatigue, pallor, shortness of breath, tachycardia and angina pectoris. Severe cases can require transfusion of red cells from blood donors. Transfusion therapy with allogeneic blood components may cause immunologic reactions and infections. In addition, repeated red blood cell transfusions can lead to iron overload. Therefore, the availability of recombinant human erythropoietin (rHu-EPO) as an anti-anemic drug has been an important medical progress.

     Initial trials of the replacement therapy with rHu-EPO to restore the hematocrit in patients with end-stage renal failure were reported about 14 years ago [1,2] which then lent a new impetus to studies of the pathophysiology and pharmacology of EPO [3]. In all likelihood, rHu-EPO is today the best selling drug in the world (estimated sales 5,000 millions US $ per year).

     This review provides basic information on: (a) the role of the hormone erythropoietin (EPO) in the control of red cell production, (b) the structural and biological characteristics of rHu-EPO preparations, (c) their clinical use in the anemia associated with renal failure and inflammatory diseases, (d) the relationship between tumor oxygenation, anemia and the efficiency of anti-tumor therapy, and (e) the potential value of the administration of rHu-EPO in the perisurgical setting including its use in autologous blood transfusion programs. Finally, (f) the problem of rHu-EPO abuse by athletes in endurance sports will be stressed. It is not intended to provide a complete survey of the relevant original publications. However, specific issues that are controversial or of prospective interest will be described in more detail.

     With respect to earlier or additional information on specific topics, the reader is referred to more complete reviews on the history of EPO research [4,5], the in vivo control of EPO synthesis [5-7], the cellular and molecular mechanisms of the action of EPO [8-11] and the clinical use of rHu-EPO in the treatment of the anemia associated with renal failure [12,13] or malignancies [14-16]. Other reviews will be cited in the text.

     Erythropoiesis counterbalances the permanent destruction of aged red blood cells by macrophages in bone marrow, spleen and liver. The basal rate of the production of red cells (2-3 x 10¹¹ per day) may 10fold increase following a blood loss. Both the basal and the augmented elaboration of red cells are mediated by the hormone EPO. Experimental studies in animals have shown that almost all circulating EPO originates from peritubular interstitial cells in the cortex of the kidneys and from parenchymal cells in the liver [17,18]. In addition, some EPO mRNA has been detected in spleen, lung, testis and brain [19-21]. Tissue hypoxia is the main stimulus of EPO production [6]. EPO levels in plasma may rise to 10,000 IU/l (International Units per liter) in severe anemia, compared to the normal value of about 15 IU/l. There is an exponential increase in EPO production with decreasing blood hemoglobin concentration in persons with intact kidneys. Hence, an inverse log/linear relationship can be drawn between the concentrations of EPO and hemoglobin in blood. This relationship is lost in patients suffering from chronic renal failure, as demonstrated in Fig. (1).

     Fig. (2) illustrates the control circuit of erythropoiesis. The O₂ capacity of the blood is the primary determinant of the rate of EPO synthesis in kidneys and liver. In addition, EPO synthesis is stimulated when the arterial O₂ tension is lowered or when the O₂ affinity of the blood is increased. Changes in renal blood flow have little influence on EPO production [22]. Several other hormones interfere with the normal pO₂-dependent feedback circuit of erythropoiesis [6]. The normally higher hemoglobin and hematocrit values in males than in females are probably due to the myeloid action of androgens, which augment the effect of EPO on erythrocytic progenitors [23,24]. Thyroid hormones may stimulate EPO gene expression [25], resulting in increased circulating EPO levels in hyperthyroidism [26].

     The molecular mechanisms of O₂ sensing are beginning to be understood [27]. Possibly, extramitochondrial heme proteins function as O₂ sensor [28-31]. A recent hypothesis suggests that H₂O₂ and other reactive O₂ species that are produced by b-type cytochromes at high pO₂ act as signaling molecules which repress EPO gene expression [32,33]. A hypoxia inducible factor (HIF-1) has been identified that binds to the hypoxia-responsive enhancer in the 3'-flanking sequence of the EPO gene. HIF-1 is a dimeric protein composed of 2 different subunits, the 120 kDa HIF-1a and the 91-94 kDa HIF-1ß [34]. HIF-1 controls the expression of several genes that encode proteins which are protective against hypoxia, such as vascular endothelial growth factor and distinct glycolytic enzymes [31,35,36].

     About 3 days after an acute increase in plasma EPO reticulocytosis becomes apparent. Reticulocytes are progenies of lineage-specific progenitors originating from a small pool of stem cells in the hemopoietic organs. Fig. (3) summarizes the main stages in erythropoiesis. The functional human EPO receptor is a 484-amino acid glycoprotein and member of the class I cytokine receptor superfamily [8,10]. The number of receptors per cell decreases with differentiation beyond the colony-forming unit-erythroid (CFU-E) level. Reticulocytes and erythrocytes have no EPO receptors.

     Fig. (4) is a schematic presentation of the EPO receptor. The first step in EPO signaling is dimerization of the receptor molecules. This induces tyrosine phosphorylation of several cytoplasmic and membrane-associated proteins including the EPO receptor itself [11]. The subsequent cellular events that lead to proliferation and differentiation have not been clearly identified. Suppression of programmed cell death (apoptosis) is the likely mechanism by which EPO maintains erythropoiesis [37]. When the concentration of the hormone rises in blood, as in anemia, an increasing number of progenitor cells escape from apoptosis and proliferate. The presence of EPO is essential for the viability, proliferation and differentiation in the erythrocytic lineage. Lack of EPO leads to anemia.

     Some evidence has been provided to assume that EPO may exert a local function in the central nervous system besides its role in erythropoiesis. EPO mRNA [19,21] and EPO receptors [38-40] can be demonstrated in brain. EPO and EPO receptor are expressed by neurons and glial cells in the brain and spinal cortex of human fetuses [41]. It is speculated that EPO may act as a neurotrophic and neuroprotective factor. Indeed, EPO has been shown to alleviate the ischemia-induced place navigation disability, cortical infarction and thalamic degeneration, when applied into the cerebrovesicles of stroke-prone rats with permanent occlusion of the left middle cerebral artery [42]. The protective effect of EPO on hippocampal cells has been also demonstrated in a gerbil forebrain ischemic model [43]. The expression of EPO as well as of the EPO receptor gene increases in response to hypoxia in brain. The cellular mechanism of the neuroprotective effect of locally produced EPO is not fully understood. It has been proposed that EPO inhibits the formation of reactive O₂  species and N-methyl-D-aspartate receptor-mediated cytotoxicity [42,43].

Production and properties of recombinant DNA-derived EPO

     In a pioneering work published in 1977 Miyake et al. [44] isolated 10 mg pure human EPO of 2550 1 urine from severely anemic patients. The preparation of pure human urinary EPO enabled it to identify the amino acid sequence of a tryptic fragment of the protein and to synthesize EPO DNA probes for the isolation and cloning of the EPO gene [45,46]. Mammalian cells transfected with the EPO gene linked to an expression vector ("recombinant DNA") produce rHu-EPO in vitro. Chinese hamster ovary (CHO) cells deficient in the dihydrofolate reductase gene are most commonly used for the large-scale pharmaceutical manufacture of the drug, because in such cells EPO gene amplification can be achieved by co-selection in the presence of methotrexate [47]. The cultures are maintained in large fermenters or roller bottles. RHu-EPO is isolated from the medium by a series of chromatography steps. Care is taken that the drug is not contaminated by microorganisms, xenogenic proteins, oncogens or pyrogens.

     Human urinary EPO and rHu-EPO are identical with regard to their amino acid sequence, position of their 2 disulfide bridges and 4 glycosylation sites, and their secondary structure. The peptide consists of 165 amino acids. The molecular mass of the glycoprotein entity is 30 kDa. The carbohydrate portion (40%) is essential for molecular stability and full in vivo biological activity. There are 3 tetraantennary N-linked (Asp 24, 38 and 83) and 1 small O-linked (Ser 126) acidic oligosaccharide side chains. The molecule forms a bundle of 4 a-helices which are folded into a compact globular structure. Although the recombinant products contain no new structure elements compared to the native hormone, there are quantitative differences in glycosylation, which may also explain the fact that the specific in vivo biological activity of purified human urinary EPO is lower (70,000 IU/mg peptide) than that of the purified recombinant product (about 200,000 IU/mg peptide).

     Like other soluble glycoproteins, native EPO exists as a pool of several isoforms that differ slightly in glycosylation. In particular, the carbohydrate side chain in position Asp 24 exhibits microheterogeneity [48]. Studies by zone electrophoresis have shown that there are intraindividual diurnal changes, interindividual differences, and abnormalities associated with inflammatory diseases with respect to the occurrence of EPO isoforms in serum [49]. Furthermore, there are differences in the electrophoretic mobility when human blood-borne, urinary and recombinant EPOs are compared [49-51]. Importantly, EPOs differing in their carbohydrate composition exhibit also differences in their biological activities and immunoreactivities [52-54]. The extent of microheterogeneity of CHO cell-expressed rHu-EPO has been studied by mass spectometry and NMR spectroscopy, as reported by Rush et al. [55].

     Two brands of CHO cell-derived EPOs, termed epoetin alfa and epoetin beta, are currently used for treatment of EPO-deficiency anemias and for support in autologous blood collection programs. Both of these types of rHu-EPO are produced in CHO cultures. In extensive studies utilising several batches of each brand Storring et al. [54] have compared epoetin alfa and epoetin beta by means of isoelectric focusing, lectin-binding, in vivo and in vitro bioassays, and immunoassays. While there were only minor inter-batch differences within the two brands, between these significant differences were apparent in the pattern of glycosylation.

Fig. (5) shows that epoetin alfa differs from epoetin beta in containing a smaller proportion of more basic isoforms. In addition, epoetin alfa possesses less N-glycans with non-sialylated outer Galb1-4GlcNAc moieties, N-glycans containing repeating Galb1-4GlcNAc sequences, tetraantennary and 2,6-branched triantennary N-glycans, and Galb1-3GalNAc containing O-glycans. Epoetin beta has a higher in-vivo: in-vitro bioactivity compared to epoetin alfa when tested in murine systems [54]. Note that this difference in biological activity cannot be explained on grounds of the degree of sialylation, because, if anything, epoetin alfa is sialylated to a greater degree than epoetin beta [54]. Clinical observations are in agreement with these results. The terminal elimination half-life of epoetin alfa is shorter than that of epoetin beta in humans following intravenous or subcutaneous administration of the drugs [56]. Still, there do not appear to be significant differences in the efficacy of the 2 brands of commercial rHu-EPO.

Attempts of developing new erythropoiesis-stimulating drugs

     According to expectations, generic rHu-EPO preparations will come to market in the near future, when the present CHO cell-derived products will no longer be protected by patent. RHu-EPO with full in vivo biological activity can be purified from the culture supernatant of other genetically engineered mammalian cells such as BHK-21 baby hamster kidney cells or C127 mouse mammary cells [57]. In addition, attempts have been successful to produce dimers and trimers of rHu-EPO which are in vivo much more efficient in stimulating erythropoiesis than the monomers [58].

     EPO gene transfer could become an alternative to the administration of rHu-EPO. Naffakh and Danos [59] have reviewed the various ex vivo and in vivo approaches for EPO gene therapy, which resulted in the production of EPO in amounts that sufficed to stimulate erythropoiesis in experimental animals. However, before clinical applications of EPO gene therapy can be explored, the stability, tissue specificity and regulatory mechanisms of the transgenes need to be more carefully investigated.

     EPO receptor agonist peptides have been isolated by random phage display peptide libraries' screening [60,61]. These EPO mimetics are cyclic 2 kDa peptides of about 20 amino acids. They show no sequence homology to EPO, but stimulate the proliferation and differentiation of erythrocytic progenitors in vitro and enhance erythropoiesis in mice in vivo. The potency of the peptides can be greatly increased through covalent dimerization [62]. The functional mimicry of a protein hormone by an unrelated small peptide is biochemically exciting. However, it is not clear whether these studies will eventually enable it to design molecules that can be administered orally or trans-dermally for stimulation of erythropoiesis in clinical practice.

Normal and abnormal levels of serum EPO

     Possible indications for assay of serum EPO in the laboratory routine include the differential diagnosis of polycythemias and anemias, the follow-up of paraneoplastic EPO production, and the election of anemic patients for rHu-EPO therapy. Radioimmunoassay or enzyme-linked immunosorbent assay procedures can be applied [63]. The sensitivity of most of the current immunoassays is still insufficient as these do not enable one to carry out valid measurements in states of low EPO concentrations, such as in polycythemia vera [64]. EPO concentrations are traditionally expressed in International Units (IU) as a measure of biological in vivo activity [65]. Reference preparations of human urinary EPO (2nd IRP) and purified DNA-derived human EPO (rDNA-derived, 130,000 IU/mg fully glycosylated protein) are available [65,66]. The concentration of serum immunoreactive EPO is 6-32 IU/1 in non-anemic individuals [63]. Interestingly, there is no significant difference detectable when the values are compared in healthy women and men, although the hemoglobin concentration is lower in the females.

     Because EPO is the only specific regulator of the growth of erythrocytic progenitors, EPO overproduction inevitably results in secondary erythrocytosis, and EPO deficiency in anemia (Table 1).

 

     Overproduction of EPO may occur as a physiological response to hypoxia, e.g. at high altitudes, or arise autonomously as a paraneoplastic syndrome. Excessive EPO production is probably the major pathogenetic factor in the development of chronic mountain sickness. Erythrocytosis increases the risk to acquire myocardial infarction and stroke. In patients suffering from secondary erythrocytosis phlebotomy treatment to hematocrit 0.52-0.50 may be beneficial to prevent hemodynamic and rheological complications such as pulmonary hypertension and peripheral thrombosis. However, repeated phlebotomies lead to iron deficiency in the long term. Unfortunately, specific erythropoiesis-inhibiting drugs have not been developed.

     Insufficient EPO production is the primary cause of the anemia in chronic renal failure [12]. The pathogenetic mechanisms may involve destruction of the EPO-producing cells, inhibition of EPO synthesis due to metabolic acidosis, accumulation of uremia toxins and increased formation of proinflammatory cytokines. Interestingly, the induction of acute renal failure in rats results in suppressed hepatic EPO gene expression [67]. Thus, the liver cannot substitute for the diseased kidneys.

     In several non-renal diseases, such as in chronic inflammation, malignancy and AIDS a relative lack of EPO contributes to the development of anemia [68,69]. Based on the results of in vitro studies, it is thought that the proinflammatory cytokines interleukin-1 (IL-1) and tumor necrosis factor-a (TNFa) suppress EPO gene expression [69]. In addition, chemotherapeutic and immunosuppressive drugs can inhibit EPO synthesis [70].

     The assessment of an "inadequate EPO response to anemia" is difficult in practice, because serum EPO levels cannot be evaluated in absolute terms but only in relation to the degree of anemia [71]. Thus, the definition of a relative deficiency of EPO relies primarily on documentation of a lowered ratio between the serum EPO level and the blood hemoglobin concentration or hematocrit in comparison with this ratio in a reference population [72]. It has been suggested to calculate the observed/predicted log [EPO] ratio (O/P ratio) for each serum sample, with the predicted level being estimated from a reference group of patients with anemia not associated with renal disease, infection, inflammation or malignancy. Such types of anemias include those of iron deficiency or hemolysis.

     Pathophysiological consequences of anemia are summarized in (Table 2).

Table 2.        Pathophysiological Consequences of Anemia

 

Pharmacokinetic of rHu-EPO

     Both intravenous and subcutaneous routes of rHu-EPO administration are effective and used in clinical practice depending on the patient's disease and accompanying medical treatment. Compared with intravenous injection, subcutaneous administration is characterized by a prolonged absorption phase (peak values reached after 12-30 h), lower peak values (about 5% of those observed after intravenous administration), lower rate of bioavailability (20-40%), but prolonged elimination half-time that has been reported in a range from 1 to 3 days [73-76]. Of note, the biological half-time of intravenously injected rHu-EPO appears to be considerably longer (4-6 h of epoetin alfa, 4-12 h of epoetin beta), when compared to the disappearance of native EPO whose half-life has been estimated to be about 2 h [77]. More detailed information on the pharmacokinetic  and pharmacodynamic of rHu-EPO is provided elsewhere [13,75,78].

RHu-EPO therapy in renal failure

     RHu-EPO as an anti-anemic drug for treatment of patients suffering from chronic renal failure was introduced 14 years ago [1,2]. Given intravenously or subcutaneously [79] it is now routinely used in patients on regular hemodialysis or continuous ambulatory peritoneal dialysis (CAPD), as well as in many predialysis patients [12,80]. At least in Europe, there appears to be a great underutilization of rHu-EPO during the predialysis period, although the correction of anemia would allow the patients to enter dialysis later than without rHu-EPO therapy and prevent left ventricular hypertrophy and congestive heart failure [81]. RHu-EPO raises hematocrit and blood hemoglobin concentration in a dose-dependent and predictable way, and it abolishes the need for red cell transfusions with its risks of incompatibility reactions, viral infections and iron overload. In previously anemic patients, rHu-EPO therapy reverses the hyperdynamic cardiac state and restores the impaired brain function. The well-being and exercise tolerance of the patients is greatly increased (Table 3).

 

 

     Eventually, rHu-EPO can correct the anemia in practically all patients with renal failure, but the dose needed is variable (Table 4). The complications responsible for rHu-EPO resistance include iron deficiency, inflammatory or infectious disease, aluminium overload, hyperpara-thyroidism, and osteitis fibrosa. In addition, the response to rHu-EPO can be improved by increasing the