Erythropoietin and iron therapy in patients with renal failure


Professor Francesco Locatelli, Department of Nephrology, Dialysis and Renal Transplantation, ‘Alessandro Manzoni’ Hospital, Via Dell' Eremo 9/11, 23900 Lecco, Italy


Anemia, which is a common complication of chronic kidney disease (CKD), may significantly impair quality of life, increase cardiovascular risk and reduce long-term survival if left untreated. Today, erythropoiesis-stimulating agents (ESAs) are the main tool for anemia correction; they can be differentiated on the basis of mean serum half life on short- and long-acting molecules, the latter requiring longer administration intervals. According to international guidelines, the target hemoglobin to be obtained by treatment is between 11 and 12 g/dL. In general, the intravenous route is more convenient for hemodialysis patients, whereas the subcutaneous one is preferable in all other CKD patients. ESA dose requirements are rarely predictable in the individual patient and thus need to be titrated according to hemoglobin increases. In order to achieve ESA effectiveness, patients often need iron supplementation, either orally or intravenously. The intravenous route is the most widely used, especially in hemodialysis patients.


Anemia is a common complication of chronic kidney disease (CKD). It is characterized by a reduced ability of the damaged kidney to produce erythropoietin (EPO), the hormone involved in proliferation and maturation of red blood cells in the bone marrow. Left untreated, anemia may significantly impair quality of life, increase cardiovascular risk and reduce long-term survival.

Previously, treatment options were essentially limited to blood transfusions; however, since the late 1980s, the availability of recombinant human erythropoietin (rHuEPO) has revolutionized the management of anemia in CKD patients. Today, erythropoiesis-stimulating agents (ESAs) are the main tool for anemia correction in CKD patients.


Anemia develops early in the course of CKD and it affects nearly all patients with CKD Stage 5; treatment with ESAs is a well-established practice, able to reduce symptoms and complications of anemia. There is full consensus that ESA therapy should be given to treat anemia to all CKD patients with a hemoglobin (Hb) level persistently below 11 g/dL after having ruled out all other causes of anemia. This applies to all CKD stages, from the early phases to patients receiving renal replacement therapy.1,2 Letting Hb levels fall too much before starting treatment exposes patients to reduced quality of life and increased morbidity. Moreover, patients starting ESA with very low Hb levels need higher ESA doses than those with milder anemia. In order to anticipate Hb decreases, anemia workup should be started before Hb levels are below 11 g/dL (< 13.5 g/dL in adult males, < 12.0 g/dL in adult females).2


A number of observational studies have described a clear relationship between anemia and mortality in CKD patients.3–5 This is probably due to the impact of chronic anemia on cardiac function, as it causes vasodilatation, cardiac dilation and increased cardiac output, leading to left ventricular dilation and compensatory hypertrophy.

Starting from this clear association, the availability of an effective therapy to treat renal anemia raised the question whether correcting anemia may improve patient outcome. Several intervention studies have been performed to test this hypothesis. Many of them were also aimed at verifying whether complete rather than partial correction of renal anemia through rHuEPO administration would lead to the best results in terms of survival or surrogate endpoints (left ventricular mass, quality of life, CKD progression). This is also important in pharmacoeconomic terms: heading to higher Hb targets implies higher ESA doses. At present, randomized clinical trials6–9 and meta-analyses10,11 do not suggest any major effect of complete anemia correction on hard, intermediate or surrogate endpoints, except for quality of life. Considering these findings, there is wide international agreement that today the most reasonable Hb target should be of 11–12 g/dL.12

The effect of Hb target on progression of kidney disease in patients not on dialysis is unclear. Individual trials showed either prolongation of kidney survival, acceleration of progression to kidney failure, or no effect, but many of the trials were underpowered to detect potentially relevant effects in either direction. For this reason, this aspect is not at present taken into account by international guidelines for the definition of the optimal Hb target.


EPO is a hydrophobic protein of 165 amino acids stabilized by three N-glycans and one O-linked sugar chain; the carbohydrate content is essential to stability and plays some important roles in the activity and biosynthesis of the molecule. The sialic acid-containing carbohydrate content of the molecule is directly related to its serum half-life and in vivo biological activity, but is inversely related to its receptor binding affinity. At present, six different types of ESAs are available on the market: epoetin alfa, epoetin beta, epoetin omega (only in Central and Eastern Europe), epoetin delta, darbepoetin alfa and continuous erythropoietin receptor activator (CERA). Mean half-life of ESAs are summarized in Table 1.

Table 1.  Mean half-life ± SD of ESA expressed in hours according to administration route
  • *

    CERA dosage of 0.4 µg/kg.

  • CERA dosage of 0.8 µg/kg. CERA, continuous erythropoiesis receptor activator; ESA, erythropoiesis-stimulating agent; ND, not determined; SD, standard deviation.

Epoetin alfa136.8 ± 2.719.4 ± 10.7
Epoetin beta138.8 ± 2.224.2 ± 11.2
Epoetin omegaNDND
Epoetin delta149.9 (SD not available)33.1 (SD not available)
Darbepoetin alfa1525.3 ± 7.348.8 ± 12.7
CERA16134 ± 19*139 ± 20

Epoetin alfa and epoetin beta are both synthesized in Chinese hamster ovary cells and share the same amino acid sequence as endogenous EPO; but differences in the manufacturing process between the two glycoproteins translate into slight differences in their carbohydrate moieties17 as well as their pharmacokinetic and pharmacodynamic properties.13 Epoetin omega is synthesized in baby hamster kidney (BHK) cells.18 It differs from epoetin alfa and epoetin beta in the proportion of O-glycosylation.19 Epoetin delta shares the same amino acid sequence as endogenous EPO, but is synthesized in human cells.20,21 This process circumvents problems arising from species-dependent differences in protein folding or post-translational modification.

Darbepoetin alfa is a hyperglycosylated EPO analog designed for prolonged survival in the circulation and thus greater biological activity. Like epoetin alfa and epoetin beta, darbepoetin alfa is produced in Chinese hamster ovary cells. Darbepoetin alfa differs from EPO in the amino acid sequence at five positions and contains five N-linked carbohydrate chains instead of three.22 As a result, it has increased molecular weight (37,100 Da compared with 30,400 Da), sialic acid content (22 compared with 14 sialic acid residues) and negative charge compared with EPO. In Sprague-Dawley rats, darbepoetin alfa given intravenously had greater in vivo efficacy than rHuEPO. This increased biological activity was due to an increase in the circulating half-life, which counterbalanced a lower relative affinity for the EPO receptor than that of rHuEPO. Based on the peptide mass, 200 IU of epoetin alfa is equivalent to 1 µg of darbepoetin alfa.

Continuous erythropoiesis receptor activator (CERA) is a new ESA, which has recently completed the Phase III program.23 It is a large molecule, approximately twice the size of EPO, which was created by integrating a single polymer chain into the EPO molecule. In vitro, CERA dissociates faster from the soluble EPO receptor than epoetin beta. It has been suggested that the binding of CERA to the EPO receptor is too brief for internalization, resulting in repeated cycles of receptor binding, stimulation and dissociation, and consequent increased erythropoietic activity.23 In 2007, the European Commission approved its use to treat anemia associated with CKD and its entering the market at the beginning of 2009.

In addition to these molecules, new agents are under clinical development. Among these, hematide is a synthetic, dimeric, pegylated peptide derived from original research on the EPO mimetics, which is undergoing Phase II of its clinical trial program.24 Its primary amino acid sequence is unrelated to that of rHuEPO. In 28 healthy male volunteers, hematide showed a dose-dependent increase in reticulocytes; the 0.1 mg/kg dose seemed to be the most effective with sustained activity for longer than 1 month.24


All ESAs are effective in correcting renal anemia and increasing Hb levels. However, ESAs differ in amino acid sequence, carbohydrate content, charge and molecular weight. These characteristics influence their half-life and biological activity and thus their clinical use. Moreover, their pharmacokinetic and pharmacodynamic properties vary according to the route of administration. This is to be chosen not only according to ESA characteristics and economical considerations, but also after taking into account CKD stage and the type of renal replacement therapy. In general, the intravenous route is more convenient for hemodialysis patients, whereas the subcutaneous one is preferable in all other CKD patients.

ESA dose requirements are rarely predictable in the individual patient and thus need to be titrated according to Hb increases. In general, predialysis patients are likely to need smaller doses than patients with CKD Stage 5. As a rule, during the correction phase, ESA requirements are 20–30% higher than during the maintenance phase. In order to avoid side effects and/or adverse events (hypertension, seizures, vascular access thrombosis), Hb should be increased slowly during the correction phase, by no more than 1–2 g/dL per month. In general, dose adjustment should not be made in the first month after the start of treatment, and not more often than every two weeks thereafter, as time is needed before significant Hb changes following dose or schedule modifications will be observed. These modifications should be determined by the rate of increase in Hb levels during the correction phase, their stability during the maintenance phase, and the frequency of Hb testing (at least monthly).2 When Hb levels exceed the target, it is warranted to decrease ESA dose, but preferably not to interrupt treatment. Indeed, this may cause Hb to decrease too much, requiring new ESA treatment at higher doses, eventually leading to excessive Hb cycling. Hb cycling has recently been cited as a risk factor for increased mortality in hemodialysis patients.25

Epoetin alfa

Epoetin alfa is administered two or three times weekly either intravenously or subcutaneously. Frequency of administration can reasonably be set at once per week in stable HD patients with low dose requirements or in predialysis patients. However, this practice is not supported by clinical trials. In 2002, following the upsurge of pure red cell aplasia (PRCA) cases, its administration by the subcutaneous route was no longer licensed for treatment of CKD patients in many countries. At present, the subcutaneous use of Eprex (Janssen-Cilag BV, Tilburg, the Netherlands) has been readmitted when the vascular access is not available in conjunction with an extensive pharmacovigilance plan. Generic formulations of epoetin alfa have been approved by the European Medicines Agency (EMEA) and are entering the EU market.

Epoetin beta

Dose requirements to maintain target Hb levels are significantly lower when epoetin beta is administered subcutaneously compared with intravenously;26 current treatment guidelines recommend the subcutaneous route of administration of epoetin beta in order to minimize treatment costs.1,2 Similarly to epoetin alfa, epoetin beta has been administered two to three times weekly. However, studies evaluating less frequent administration regimens have demonstrated that once weekly subcutaneous administration during the maintenance phase has the same efficacy as the three-times-weekly regimen.27,28

Epoetin delta

The clinical experience with this agent is still limited, as it has only recently been put on the market. Data from clinical trials indicate that it is effective in correcting renal anemia in rHuEPO-naive patients.21,29 Its efficacy and dose pattern are similar to those of the other two rHuEPOs.

Darbepoetin alfa

Given its longer half-life than rHuEPO, darbepoetin alfa can be administered once a week30–32 or once every other week.33 According to the label, the drug can be administered also once a month. In dialysis patients, this may require higher doses to achieve a given Hb target compared with more frequent administration. Data from secondary analyses31,32 and from one prospective, randomized crossover study34 suggest that dose requirements are independent of the administration route. In other words, patients given the drug intravenously need the same dose as that given subcutaneously.

Continuous erythropoietin receptor activator

Phase II35,36 and III37,38 studies indicate that CERA corrects anemia and maintains Hb levels within guideline targets when administered up to once monthly in predialysis and dialysis patients. The most suitable starting dose seems to be 0.60 µg/kg given twice monthly.


A clear relationship among Hb levels, ESA dose and increase in dialysis dose has been pointed out by a number of prospective or retrospective studies.39,40 This is particularly true in patients receiving inadequate dialysis.39 Increasing attention has also been paid to the relationship between dialysis, increased inflammatory stimulus and ESA response, as dialysate contamination and low-compatible treatments may increase cytokine production and consequently inhibit erythropoiesis. The biocompatibility of dialysis membranes and flux are other important factors. However, in highly selected, adequately dialyzed patients without iron or vitamin depletion, the effect of these treatment modalities on anemia seems to be smaller than expected.41 The role of online treatments is still controversial, given that it is still difficult to discriminate between the effect of online hemodiafiltration per se from that of an increased dialysis dose.42


Dose requirements to achieve anemia correction are quite variable and poorly predictable in the individual patient. However, a number of patients need a greater-than-usual ESA dose and are defined as hyporesponsives. According to the last revision of the European Best Practice Guidelines (EBPG),1 resistance to ESA treatment is defined as a continued need for > 20,000 IU/week (300 IU/kg/week) of rHuEPO administered subcutaneously or 1.5 µg/kg of darbepoetin alfa (greater than 100 µg/week); this means that resistant patients require more than 2.5 times the average ESA dose. The true incidence of ESA hyporesponsiveness is still a matter of study, and probably differs from country to country (in the United States it is likely to be higher than in Europe). According to the definition above, the prevalence of resistance to ESA was only of 2.4% in a recent cross-sectional study of 550 Italian hemodialysis patients.43 Conversely, the prevalence of hyporesponsiveness to ESA was much greater in a cohort from a large dialysis organization (DaVita) in the United States (43% of the patients received more than 18,000 IU/week of rHuEPO).25 In this population of nearly 60,000 subjects, requiring higher ESA doses was a marker of higher death risk. Hyporesponsiveness also occurs in predialysis patients.44

Causes of incomplete response to ESAs are summarized in Table 2. The most common one is iron deficiency – absolute or functional. According to an Italian cross-sectional study,43 16% of the patients had a transferrin saturation of less than 15%, which is considerably below that recommended in the EBPG and the National Kidney Foundation-Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines. Occult blood loss, infection, inflammation and inadequate dialysis are also important causes. In recent years, increasing attention has been paid to the relationship between dialysis, increased inflammatory stimulus, malnutrition and ESA response.43,45 Angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor antagonists may also play a role. Compliance should be checked in patients self-administering an ESA.

Table 2.  Main causes of resistance to treatment with ESA
 Chronic kidney disease
  1. ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; ESA, erythropoiesis-stimulating agent.

Iron deficiency
Chronic blood loss
Chronic infections and inflammation 
Hyperparathyroidism/osteitis fibrosa 
Aluminum toxicity 
Multiple myeloma, myelofibrosis 
Vitamin deficiencies (e.g. folate or vitamin B12)
Dialysis-related carnitine deficiency 
Inadequate dialysis 
Cytotoxic and immunosuppressive agents 
ACE inhibitors or ARBs 
Pure red cell aplasia 


Antibody-mediated PRCA is a serious adverse event related to ESA therapy. In this disease, epoetin-induced antibodies neutralize all the exogenous rHuEPO and cross-react with endogenous EPO. As a result, serum EPO levels are undetectable and erythropoiesis becomes ineffective. Despite the widespread use of rHuEPO, PRCA remained a very rare complication for many years. Since 1998, the number of reported cases has increased dramatically;46 the majority of the cases were observed in patients treated with subcutaneous Eprex, the epoetin alfa produced outside the United States. The upsurge coincided with the substitution of human serum albumin by polysorbate 80 in the Eprex formulation. Polysorbate 80 may elicit the formation of epoetin-containing micelles that could be immunogenic. Alternatively, leachates released by uncoated rubber stoppers of prefilled syringes may interact with polysorbate 80 and act as an adjuvant of the immune reaction. From December 2002, the subcutaneous use of Eprex in CKD patients was contraindicated in Europe by regulatory authorities and was strongly discouraged in Canada and Australia. Starting from 2003, the number of reported cases dramatically dropped. This may have been caused by the shift in administration route, reinforcement of product cold chain, or elimination of uncoated rubber syringe stoppers. Interestingly, despite of the large use of rHuEPO in oncology, cases were identified only in CKD patients.

The number of reported cases of PRCA has decreased sharply since 2003 and with no more cases reported in 2007. This may be due to a change in the route of administration, the reinforcement of the product cold chain or the elimination of uncoated rubber syringe stoppers.


Despite widespread use, ESAs are effective and safe products. Nevertheless, a number of complications are described (Table 3).

Table 3.  Most common side effects of ESA therapy
  1. ESA, erythropoiesis-stimulating agent.

Diabetic retinopathy
Thrombotic complications
Vascular access thrombosis
Metabolic disturbances
Injection-site pain
Pure red cell aplasia


A number of pathophysiological mechanisms may explain ESA-induced rise in blood pressure.47 The increase in blood viscosity secondary to anemia correction appears to be the most obvious one. This is particularly true when anemia correction is achieved too rapidly or higher Hb targets are reached.11 However, blood-pressure changes are often not clearly related to achieved Hb levels. Moreover, even single dose administrations of ESA are capable of inducing hypertension in some patients. Enhanced vascular reactivity and vasoconstrictor responses have been suggested to play a role.


Seizure was first reported in early clinical trials in patients who developed severe hypertension in association with a rapid increase in hematocrit.48 Nowadays, this side effect is quite rare.

Thrombotic complications

An association of ESA therapy with vascular thrombotic events has been suggested. This complication is more likely when Hb is increased too fast or when it largely exceeds the target, especially in patients with diabetes or already established cardiovascular disease.6,7 According to a recent meta-analysis of nine clinical trials,11 the risk of arteriovenous access thrombosis is significantly higher in patients randomized to near-to-normal Hb levels than in patients randomized to a lower target.


Clinical trials have noted a 15–17% frequency of headaches in patients receiving ESA.49 However, the role of ESA is unclear, as end-stage renal disease patients not receiving ESA have a similar rate for headache.49 Headache is generally mild and usually does not preclude treatment.

Diabetic retinopathy

EPO increases proliferation of vascular endothelial cells and is a potential retinal angiogenic factor. Experimental data suggest that the binding of EPO to its receptor leads to activation of the mitogen-activated protein (MAP) kinase pathway; this pathway may elicit angiogenesis in diabetic retinopathy.50 At present, there is no clinical evidence indicating that ESA is the cause of proliferative retinopathy.


In CKD patients, iron therapy is not only aimed at correcting iron deficiency, but is also an adjuvant therapy in patients receiving ESA to achieve and maintain the Hb target. In these patients, iron stores may be nearly normal, but during ESA treatment, there may be insufficient immediately available iron to optimize ESA therapy. In this context, iron therapy significantly reduces ESA dose requirements.

Iron status

According to clinical needs, iron status testing should be made every 1–3 months.1,2 This information should then be weighted together with Hb level, ESA dose and their trend over time, in order to elucidate the status of both external iron balance (gain or losses) and internal iron balance (distribution of iron in stores and erythrocytes).2

Traditional and more widely used iron tests are serum ferritin and transferrin saturation (TSAT) levels. However, these are not optimal tests, as they lack accuracy and stability. Indeed, they are influenced greatly by inflammation and malnutrition, two conditions often affecting CKD patients. For this reason, there has been interest in developing other iron status tests for use in patients with CKD. Two of these, the percentage of hypochromic red blood cells (%HRBC) and the reticulocyte hemoglobin content (CHr), are the most reliable, providing direct insight into bone marrow iron supply and utilization. According to the data by Tessitore et al.,51 a %HRBC level greater than 6% is the single most accurate predictor of response to intravenous iron treatment in hemodialysis patients. Unfortunately, %HRBC is affected by inflammation and is positively influenced by erythropoietic activity, as reticulocytes are considered hypochromic by cell counters. CHr has been found to be an early predictor of response to iron therapy in hemodialysis patients.52 However, the cut-off of this marker to discriminate iron deficiency still needs to be fully clarified.53 At present, neither test is as easy to use, cost-effective and widely available as the traditional tests, such as serum ferritin and TSAT.

The reticulocyte hemoglobin equivalent (RET-He), recently introduced to determine the forward scatter of fluorescence-labeled reticulocytes, seems to be a sensitive indicator of iron-deficiency anemia. Compared with CHr, the value of 30.5 pg for RET-He appears to be the best cut-off point with a very good sensitivity and specificity to determine patients needing iron supplementation.54 Combined use of CHr and high-fluorescence reticulocyte count is very accurate in predicting response to intravenous iron therapy in hemodialysis patients.55

Soluble transferrin receptor (sTfR) is not affected by acute inflammation; however, it reflects ongoing erythropoiesis and not iron availability.56 Zinc protoporphyrin (ZPP) concentration has also been suggested as an indicator of functional iron deficiency. In cases of iron deficiency, zinc replaces iron in newly formed protoporphyrin IX to form ZPP. However, ZPP is an inferior measure of iron availability and ESA response compared with %HRBC and CHr.1

An ideal marker of functional iron deficiency should be independent of erythropoietic activity. New cell counters are able to determine cell volume and Hb concentration separately on reticulocytes and mature erythrocytes. According to Bovy et al.,57 overall RBC was not significantly different from RBC assessed only in mature erythrocytes.

Targets of iron therapy

For patients on hemodialysis, the last available international guideline on anemia2 recommend the following iron targets in hemodialysis patients:

  • serum ferritin: 200–500 ng/mL
  • TSAT: > 20% or CHr > 29 pg/cell

Support for the guideline comes from several interventional trials.58–61 The upper limit of serum ferritin of 500 ng/mL was chosen in order to minimize the risks of iron overload, without denying iron therapy to inflamed patients who may have high ferritin levels but functional iron deficit.

Evidence for iron target in CKD patients not on dialysis and in patients on peritoneal dialysis (PD) is poorer. In these patients, lower serum ferritin levels are probably adequate to ensure effective erythropoiesis with ESA treatment, with a suggested target of 100–500 ng/mL.2

Iron administration

There is wide consensus that the preferred route of iron administration is intravenous in hemodialysis patients; in PD patients and CKD patients not on dialysis, the route of iron administration can be either intravenous or oral.1,2

Oral iron is absorbed best when given without food; constipation, diarrhea, nausea or abdominal pain limit compliance.

Iron dextran, ferric gluconate and iron sucrose are available for intravenous administration. They differ in pharmacokinetics, maximum dose size, maximum rate of infusion,62 and in the rate of adverse reactions.63,64 In particular, anaphylactic reactions have been mainly described following administration of iron dextran.65 This is particularly true for the high-molecular weight formulation; ferric gluconate and iron sucrose are associated with lower rates of serious adverse events.66,67 The safety and the efficacy of ferumoxytol, a semisynthetic carbohydrate-coated iron oxide, are undergoing phase III clinical trials.68

There are two main approaches to IV iron administration, both valuable and widely used. The first is the episodic administration of a series of IV doses when iron tests go below the target; the second is the regular administration of smaller doses to maintain stable iron levels.69 No randomized clinical trial so far has compared the efficacy and safety of these two approaches.

Some worries persist on possible long-term complications of IV iron therapy. An increased risk of infection has been suggested to be associated with iron overload and iron administration.70 In addition, iron therapy may cause in vitro and in vivo oxidation of lipids and proteins, leading to oxidative damage.71 Their clinical relevance is still to be elucidated.