Erythropoiesis-stimulating agent hyporesponsiveness (Review Article)


Professor David W Johnson, Department of Renal Medicine, Level 2, Ambulatory Renal and Transplant Services Building, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Brisbane, Qld. 4102, Australia. Email:


SUMMARY:  Approximately 5–10% of patients with chronic kidney disease demonstrate hyporesponsiveness to erythropoiesis-stimulating agents (ESA), defined as a continued need for greater than 300 IU/kg per week erythropoietin or 1.5 μg/kg per week darbepoetin administered by the subcutaneous route. Such hyporesponsiveness contributes significantly to morbidity, mortality and health-care economic burden in chronic kidney disease and represents an important diagnostic and management challenge. The commonest causes of ESA resistance are non-compliance, absolute or functional iron deficiency and inflammation. It is widely accepted that maintaining adequate iron stores, ideally by administering iron parenterally, is the most important strategy for reducing the requirements for, and enhancing the efficacy of ESA. There have been recent epidemiologic studies linking parenteral iron therapy to an increased risk of infection and atherosclerosis, although other investigations have refuted this. Inflammatory ESA hyporesponsiveness has been reported to be improved by a number of interventions, including the use of biocompatible membranes, ultrapure dialysate, transplant nephrectomy, ascorbic acid therapy, vitamin E supplementation, statins and oxpentifylline administration. Other variably well-established causes of ESA hyporesponsiveness include inadequate dialysis, hyperparathyroidism, nutrient deficiencies (vitamin B12, folate, vitamin C, carnitine), angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, aluminium overload, antibody-mediated pure red cell aplasia, primary bone marrow disorders, myelosuppressive agents, haemoglobinopathies, haemolysis and hypersplenism. This paper reviews the causes of ESA hyporesponsiveness and the clinical evidence for proposed therapeutic interventions. A practical algorithm for approaching the investigation and management of patients with ESA hyporesponsiveness is also provided.

The development of erythropoiesis-stimulating agents (ESA), such as recombinant human erythropoietin (EPO) and darbepoetin alfa (DPO), has resulted in substantial health benefits for patients with end-stage renal failure (ESRF), including improved quality of life, reduced blood transfusion requirements, decreased left ventricular mass, diminished sleep disturbance and enhanced exercise capacity.1,2 Unfortunately, a considerable proportion of such patients exhibit a suboptimal haematologic response to ESA, as evidenced by the persistence of anaemia despite adequate dosing, or the requirement of high doses to achieve recommended haemoglobin targets.3 The Kidney Disease Outcomes Quality Initiative (KDOQI) Guidelines and the European Best Practice Guidelines define ESA hyporesponsiveness as a continued need for greater than 300 IU/kg per week EPO or 1.5 μg/kg per week DPO administered by the subcutaneous route.4,5 For intravenous (i.v.) ESA administration, the threshold for diagnosing EPO resistance is raised to 4004 or 450 IU/kg per week.5 This paper reviews the common causes of ESA hyporesponsiveness and recommends an approach to their investigation and management.


The most common cause of ESA hyporesponsiveness is absolute iron deficiency (defined as a ferritin concentration less than 100 μg/L with or without reduced transferrin saturation (TSAT) levels) or functional iron deficiency (defined as a ferritin concentration greater than 100 μg/L associated with a TSAT <20%).4,5 According to the European Survey on Anaemia Management (ESAM) 2003,6 34% of ESA-treated patients failed to achieve a haemoglobin level of 0e;11 g/dL and 51.6% were assessed as having inadequate iron status, defined as a serum ferritin concentration <100 μg/L and/or a TSAT <20% (or a hypochromic red cell value of >10%). Adequate i.v. iron therapy has now been widely acknowledged as an important strategy for optimizing the haematologic response to ESA and numerous controlled trials have demonstrated the superiority of i.v. over oral iron supplementation in haemodialysis,7 PD8 and predialysis patients.9 In spite of this evidence, a recent US survey demonstrated that i.v. iron supplementation was used in less than 55% of haemodialysis patients and less than 10% of peritoneal dialysis (PD) patients treated with ESA.10 The ESAM reported similar findings,11 in which 60% of patients with identified absolute iron deficiency in the first month of the study had still not received iron supplementation by the end of the study. Haemodialysis patients were more likely to be iron-deficient and require iron supplementation than PD patients, presumably related to greater dialysis-related blood losses in the former group (by up to 2.5 L or 1000 mg iron per year). Iron supplementation has also been shown to improve ESA responsiveness even in patients with adequate iron stores,12 although this strategy appears to lose its effectiveness once serum ferritin concentrations exceed 500 μg/L.13 De Vita et al. demonstrated that EPO- and i.v. iron-treated haemodialysis patients achieved a higher haemoglobin at a lower EPO dose when the target serum ferritin was 400 ng/mL as compared with 200 ng/mL.14 Similarly, Besarab et al. showed a 40% reduction in EPO dose with TSAT targets of 30–50% compared with 20–30%.15 The Caring for Australasians with Renal Insufficiency (CARI) guidelines16 recommend optimization of ESA dose by achieving higher target values for iron storage that are well above the levels that define absolute or relative iron deficiency (i.e. serum ferritin 200–500 ng/mL and TSAT 30–40%). In contrast, the KDOQI Guidelines are more conservative and recommend maintaining a serum ferritin value above 200 ng/mL in haemodialysis patients and 100 ng/mL in other chronic kidney disease (CKD) patients, a maximum ferritin level of 500 ng/mL and a minimum TSAT value of 20%.5

There is ongoing controversy as to whether giving iron to patients with a TSAT <20% and a ferritin >500 ng/mL is futile.17 The commonly held interpretation is that a low TSAT and high ferritin represents an acute-phase response to inflammation, a situation in which further iron administration is unlikely to be helpful and may be potentially harmful as a result of the pro-inflammatory effects of i.v. iron.17 Published outcomes of iron supplementation in this setting have been conflicting,15,18 as have attempts to better characterize iron status with reticulocyte haemoglobin content and serum soluble transferrin receptor.17 Coyne et al. recently published the results of the Dialysis patients' Response to IV iron with Elevated ferritin (DRIVE) trial, in which 134 patients with haemoglobin 0e;110 g/L, serum ferritin 500–1200 ng/mL, TSAT 0e;25% and epoetin dosage 0e;225 IU/kg per week or 0e;22 500 IU per week were randomly assigned to receive either no iron (control) or ferric gluconate 125 mg intravenously with eight consecutive haemodialysis sessions.19 Epoetin dosage was increased by 25% in both groups at randomization. The subsequent increase in haemoglobin level at 6 weeks was significantly higher in the i.v. iron group compared with controls (16 ± 13 vs 11 ± 14 g/L, P = 0.028). No safety issues were identified, although the trial was neither powered for a safety assessment nor sufficiently long to assess medium-to-long-term safety. Based on the available evidence to date, serum ferritin and TSAT still remain the most clinically useful parameters for initiating and monitoring iron therapy. In the setting of a low TSAT and high ferritin level, increasing ESA dosage is warranted. If haemoglobin targets have not been attained, it may be reasonable to administer i.v. iron even if the serum ferritin concentration exceeds 500 ng/mL, although the long-term safety of this practice is uncertain.

There is no evidence to guide whether i.v. iron is best delivered as small, frequent boluses or larger, less frequent boluses, although a small randomized controlled trial found no difference in haematologic response between these different regimens.20 Administration of i.v. iron may occasionally induce acute ‘free iron’ reactions, characterized by hypotension, dyspnoea, arthralgia, myalgia, nausea, vomiting and abdominal or back pain. These are dose- and rate-dependent and are rarely seen with dosages of 300 mg or less.21,22 Idiosyncratic anaphylactic reactions may also occur, but have been almost exclusively reported in relation to iron dextran (0.7% incidence).23,24 Recent epidemiological data have linked augmented body iron stores with increased risks of both cardiovascular disease25,26 and bacterial infections,27 although this evidence has been refuted by other investigations.28,29 In a 6-month surveillance study (EPIBACDIAL) of 998 haemodialysis patients, anaemia, but not i.v. iron therapy, was associated with an increased risk of experiencing at least one bacteraemic episode.28 Kalantar-Zadeh et al. observed that i.v. iron doses in excess of 400 mg/month were associated with an increased risk of all-cause and cardiovascular death in a retrospective, observational cohort study of 58 058 haemodialysis patients.30 Feldman et al. similarly reported a higher mortality rate in ESRF patients prescribed more than 1000 mg of iron dextran over a 6 month period,31 although a subsequent analysis by the same author found no statistically significant association between any level of iron administration and mortality after fitting multivariable models that appropriately accounted for time-varying measures of iron administration, as well as other fixed and time-varying measures of morbidity.29 They concluded that the previously observed association between iron administration and higher mortality was likely confounded by incomplete representation of iron dosing and morbidity over time. Based on the available evidence to date, both the CARI16 and KDOQI Guidelines5 recommend the i.v. route as the preferred route of administration of supplemental iron, particularly in haemodialysis patients. It is also preferable from an evidence-based perspective to use the i.v. route in PD and predialysis patients, if logistically feasible for the renal unit. Iron stores (TSAT and ferritin) should be monitored three monthly, with measurements deferred for 1 week following an i.v. iron dose of <200 mg iron polymaltose, or for 2 weeks if larger dosages are used.16


Inflammation is being recognized with increasing frequency as a cause of ESA hyporesponsiveness. Even in the absence of acute infection or other clearly identifiable causes of systemic inflammation, abnormally elevated C-reactive protein (CRP) levels have been observed in 30–50% of predialysis,32,33 PD34 and haemodialysis patients.35,36 Several studies have clearly demonstrated an inverse association between inflammatory indices and ESA responsiveness.11,37–40 The ESAM reported that weekly EPO doses were 80% higher in patients with serum CRP concentrations 0e;20 mg/L compared with those whose levels were <20 mg/L.11 Moreover, EPO responsiveness continued to worsen over the 6 month observation period in the high CRP group. The causes of inflammation in uraemia are poorly understood, but are likely to include decreased renal clearance of pro-inflammatory cytokines (including tumour necrosis factor-α, interleukin-1, interleukin-6 and interferon-γ), and increased cytokine production as a result of comorbidity (such as heart failure, atherosclerosis and volume overload), accumulation of advanced glycation end-products, carbonyl stress, oxidative stress, unrecognized persistent infections (such as Chlamydia pneumoniae), periodontal infections, drugs (such as iron), haemodialysis catheters, arteriovenous grafts, failed transplant allografts, bioincompatible haemodialysis membranes, bioincompatible dialysates, use of non-sterile dialysates, and/or PD-associated peritonitis.3241–45 These inflammatory cytokines are in turn thought to directly inhibit erythropoiesis and promote apoptosis of erythroid precursors.46,47 Furthermore, CKD has been associated with enhanced hepatocyte synthesis of hepcidin, which in turn promotes functional iron deficiency via inhibition of the release of iron to erythroid progenitors from reticuloendothelial cells.48

Inflammatory ESA hyporesponsiveness has been reported to be improved by a number of anti-inflammatory interventions, including use of biocompatible membranes,39 ultrapure dialysate,49,50 transplant nephrectomy,51 ascorbic acid therapy,52 vitamin E supplementation,53 statins54 and oxpentifylline administration.55,56 Uncontrolled studies have suggested that the use of high-flux biocompatible membranes may be associated with enhanced EPO responsiveness,57–61 possibly as a result of augmented removal of medium- to large-molecular-weight erythroid inhibitors.62,63 However, a subsequent multicentre, randomized controlled trial of high-flux synthetic versus low-flux cellulose membranes in 84 haemodialysis patients with ESA-hyporesponsive anaemia (representing 5% of the total haemodialysis population) failed to observe any significant differences in either haemoglobin levels or ESA dosages.60 Comparable findings were reported by a randomized cross-over trial of high- versus low-flux polysulphone dialysers in 25 haemodialysis patients.64 Two small (n = 30 and 34) randomized controlled trials have demonstrated that patients treated with online-produced ultrapure dialysate exhibited significant reductions in bacterial dialysate contamination, circulating inflammatory markers (CRP and interleukin-6) and ESA doses.49,50 Similarly, an open-label randomized controlled trial of i.v. ascorbic acid (300 mg each dialysis session) in 46 patients with refractory anaemia (EPO dose 0e;450 IU/kg per week and serum ferritin 0e;500 μg/L) reported that the antioxidant intervention was associated with significant increases in haemoglobin and TSAT levels and decreases in EPO doses and serum CRP concentrations.52 These findings are supported by those of a double-blind randomized cross-over trial of i.v. ascorbic acid therapy in 63 ESA-treated (but not necessarily ESA-hyporesponsive) haemodialysis patients,65 although a recently published, small (and probably underpowered) randomized controlled trial observed no significant effect of either oral or i.v. ascorbic acid (250 mg thrice weekly) on haemoglobin levels or iron availability in iron-overloaded haemodialysis patients.66 Treating haemodialysis patients with vitamin E-bonded dialysers for 1 year has also been shown in a small randomized controlled trial to be associated with significant reductions in EPO dose from 5383 ± 2655 U/week to 4235 ± 3103 U/week.53 The antioxidant effects of vitamin E may be additive with those of vitamin C.67 Statin administration has also been reported to promote anti-inflammatory actions, which in one retrospective study was associated with an 18% increase in haemoglobin levels and a 25% decrease in ESA requirements.54 Oxpentifylline has been found to exhibit important anti-inflammatory properties, which may be beneficial in ESA-hyporesponsive patients. Navarro et al. treated seven anaemic patients with advanced kidney disease (creatinine clearance <30 mL/min) with oxpentifylline (400 mg daily per os) for 6 months.56 Haemoglobin levels significantly increased from 99 ± 5 to 106 ± 6 g/L (P < 0.01), while serum TNF-α concentrations decreased from 623 ± 366 to 562 ± 358 pg/mL (P < 0.01). No changes were observed in untreated controls. Similarly, Cooper et al. administered oxpentifylline (400 mg daily per os) for 4 months to 16 ESRF patients with EPO-resistant anaemia (defined as a haemoglobin level <107 g/L for 6 months before treatment and an EPO dose of 0e;12 000 IU/week).55 Among the 12 patients who completed the study, mean haemoglobin concentration increased from 95 ± 9 to 117 ± 10 g/L (P = 0.0001), whilst ex vivo T cell generation of TNF-α and IFN-γ was significantly reduced. A multicentre, randomized controlled trial of oxpentifylline versus placebo in 110 ESA-resistant patients is currently underway in Australia, New Zealand and the United Kingdom.

At present, there is not a sufficient body of high-quality, consistent evidence to recommend any particular intervention for inflammatory anaemia. Some relatively inexpensive therapies, such as vitamin C, are associated with conflicting experimental observations and can lead to secondary oxalosis (particularly if serum oxalate levels are not monitored and ascorbic acid doses are not titrated to maintain serum oxalate levels below 50–100 μmol/L). If no other sources of inflammation are identified, it is probably reasonable to consider transplant nephrectomy (especially if the residual renal function is minimal), statins (especially hypercholesterolaemic patients) and oxpentifylline (especially patients with peripheral vascular disease for which condition the drug is registered). The additional expense associated with high-flux membranes (approximately $2500 per annum in Australia) and ultrapure dialysate to treat ESA hyporesponsiveness is probably difficult to justify at this stage based on the available evidence and lack of cost-effectiveness studies.


Erythropoiesis-stimulating agents non-compliance, defined as less than 90% use of the prescribed dose, has been identified in 35–55% of the dialysis population,68,69 and is a common cause of ESA hyporesponsiveness. The reported significant independent predictors of non-compliance include self-administration of ESA, younger age, higher education level and non-compliance with peritoneal dialysis exchanges. In one study, 58% of non-compliant patients were identified by direct questioning, while 74% were identified by pharmacy record review.69 Forgetfulness was the most commonly cited reason for non-compliance followed by injection pain. Simplified dosing regimens, such as once weekly dosing, may improve compliance, although studies have yet to examine compliance rates with the use of once-weekly versus once-fortnightly versus less frequent dosing schedules. Avoiding self-administration may also be desirable in some circumstances. If injection pain is an issue, trying an alternative ESA may be helpful. Head-to-head trials of injection pain associated with different ESA preparations are currently underway. Finally providing tailored education and discussion of improved outcomes associated with ESA have also been advocated as a means of enhancing compliance.70 Although these suggestions intuitively would seem likely to affect compliance, there have not been any interventional studies published in this area to date.


Erythropoiesis-stimulating agents responsiveness has been shown to be positively associated with both dialysis duration71 and delivered dialysis dose, as determined by urea reduction ratio72 or Kt/V.59,73 In a prospective study of 20 consecutive haemodialysis patients, raising the mean urea reduction ratio from 60.7% to 72% for 6 weeks was accompanied by a rise in haematocrit from 28.4 ± 0.78% to 32.3 ± 0.71% (P = 0.002).72 In contrast, mean haematocrit did not change in a control group of 20 patients with equivalent baseline urea reduction values in whom the dialysis level was not altered (28.2 ± 0.84% to 26.3 ± 0.85%; P = 0.175).72 Preliminary, small, pretest/post-test studies have also suggested that enhanced frequency (short daily or nocturnal) haemodialysis may reduce ESA requirements by up to 39%.74 However, a small (n = 84), short-duration (12 weeks) randomized controlled trial was unable to demonstrate a benefit of high-flux haemodialysis on ESA responsiveness.60


Secondary hyperparathyroidism has frequently been implicated in ESA hyporesponsiveness, possibly as a result of increased red blood cell fragility,75 direct inhibitory effects of parathyroid hormone on EPO synthesis and erythroid progenitors and an indirect effect via bone marrow fibrosis.46,76 Mandolfo et al. observed a 20% increase in haemoglobin level despite a 34% decrease in EPO dosage in 19 dialysis patients following parathyroidectomy.77 Lee et al. reported comparable results in 32 haemodialysis patients undergoing parathyroidectomy.78 Similarly, Neves et al. reported a statistically significant 12% increase in haemoglobin level without a significant change in ESA dosage in 11 elderly CKD patients with secondary hyperparathyroidism treated for 12 months with regular i.v. calcitriol.79 There have been no published studies to date concerning the impact of cincalcet on anaemia associated with secondary hyperparathyroidism.


Erythropoiesis-stimulating agents hyporesponsiveness has infrequently been attributed to vitamin B12 deficiency80 or folate deficiency.81–83 Vitamin C deficiency is also associated with decreased iron availability, which may be improved by ascorbic acid supplementation.84 Carnitine deficiency has been more frequently implicated as a causative factor in ESA hyporesponsiveness.85–87 However, the evidence that carnitine supplementation enhances ESA responsiveness has been conflicting.88 A recent systematic review of six small randomized controlled trials89 found a significant benefit of carnitine supplementation in ESA-treated patients, although the validity of this result is questionable given the presence of significant trial heterogeneity. In spite of the paucity of available evidence, the KDOQI Guidelines recommend an empirical trial of carnitine supplementation for hyporesponsive ESA-dependent anaemia.5,90


Angiotensin-converting enzyme inhibitors (ACEi) and angiotensin receptor blockers (ARB) have been purported to promote ESA hyporesponsiveness via numerous mechanisms, including inhibition of angiotensin II-induced EPO release and augmentation of plasma levels of N-acetyl-seryl-aspartyl-lysyl-proline (which prevents the recruitment of pluripotent haemopoietic stem cells).76,91 However, there have been at least 16 observational cohort studies, with approximately half demonstrating a neutral effect and half demonstrating a deleterious effect of ACEi and ARB on ESA responsiveness (reviewed in Locatelli et al.4). Nevertheless, it seems prudent to trial cessation of ACEi/ARB in patients if there is no other clearly definable cause for their ESA hyporesponsiveneness, and if the anticipated benefits exceed the anticipated risks.


Aluminium overload has rarely been reported to cause a microcytic anaemia that is poorly responsive to ESA, because of interference with the enzymes involved in heme synthesis.92 It is often difficult to ascertain if patients have aluminium overload since serum aluminium concentrations correlate poorly with the presence of bone aluminium deposition.93 The KDOQI Guidelines recommend that a desferrioxamine infusion test94 be performed in patients suspected of having aluminium overload or toxicity, especially in the setting of high baseline serum aluminium levels (>2.2 μmol/L). Chelation therapy with desferrioxamine has been shown to improve ESA responsiveness in a limited number of patients,95,96 although such therapy results in a high incidence of side-effects, especially visual toxicity.97


Patients with cancer frequently demonstrate hyporesponsiveness to ESA (mean 40%, range 15–75%).98,99 ESA hyporesponsiveness is most commonly observed in myelodysplastic syndromes and least frequently seen in multiple myeloma and chronic lymphocytic leukaemia.99 Resistance to treatment can often be overcome by the prescription of larger ESA doses.100 Published international guidelines for ESA administration in multiple myeloma and chronic lymphocytic leukaemia101 recommend initiating EPO at 10 000 IU thrice weekly or 40 000 IU weekly to maintain haemoglobin concentration at 100 g/L, only after other possible causes of anaemia have been eliminated. Non-responsive patients (<10 g/L over 4 weeks) may have their dose increased to 20 000 IU thrice weekly or 60 000 IU weekly. The ESA should be discontinued if there is no response to this increased dosage or if the haemoglobin level exceeds 140 g/L. Interestingly, the use of ESA in patients with multiple myeloma has been associated with increased overall survival, possibly related to direct immunologic and antineoplastic actions of ESA.102 It should also be noted that the US Food and Drug Administration has recently issued an alert following analyses of four new randomized controlled studies in patients with cancer which found a higher chance of serious and life-threatening side-effects (including tumour progression) and/or death with the use of ESA ( The use of ESA is therefore not approved to treat anaemia in cancer patients not receiving chemotherapy as it offers no proven benefits under such circumstances and may shorten the times to tumour progression, disease recurrence and death.


ESA-induced, antibody-mediated pure red cell aplasia (PRCA) is a rare haematological condition characterized by severe, transfusion-dependent, regenerative anaemia because of the production of neutralizing anti-ESA antibodies. While antibody-mediated PRCA was extremely rare before 1998, the incidence of this disorder increased sharply after this time in patients receiving subcutaneous epoetin alfa produced by Ortho-Biotech and marketed outside the United States.103 The level of antibody-mediated PRCA associated with epoetin alfa administration seems now to have returned to baseline levels (approximately 0e;1 case per 10 000 patient-years of ESA exposure) following increased attention to proper product handling and storage, reduced use of the subcutaneous route, and the use of Teflon-coated rubber stoppers in pre-filled syringes. In typical cases, patients initially respond to ESA treatment, followed by a drop in haemoglobin level at the rate of about 1 g/L per day without transfusions (equivalent to a transfusion requirement of approximately 1 unit of packed red blood cells per week to maintain haemoglobin level) and a sudden decrease in blood reticulocyte count to less than 15 × 109/L with normal platelet and white cell counts.103 The diagnosis of antibody-mediated PRCA relies mostly on the results of bone marrow biopsy or aspirate, which shows an absence of erythroid precursors and/or red cell maturation arrest while counts of white cell and platelet precursors are normal, and on the identification of circulating anti-EPO antibodies. In many patients, ESA cessation is coupled with immunosuppressive drug treatment to lower the level of anti-EPO antibodies. Therapy with corticosteroids plus cyclophosphamide, cyclosporine A alone, or renal transplantation has been shown to result in the best recovery rates, while simple EPO withdrawal or i.v. immunoglobulin alone yielded little or no recovery from PRCA.104 Another promising advance is the recent development of EPO-mimetic peptides, such as hematide, which have been shown to stimulate erythropoiesis and increase haemoglobin in animals with circulating anti-EPO antibodies,105 suggesting that this compound could potentially be useful as rescue therapy in patients with antibody-mediated PRCA.


Other potentially modifiable causes of ESA hyporesponsiveness include primary bone marrow disorders (e.g. myelodysplasia, myelofibrosis), myelosuppressive agents (e.g. cytotoxic and immunosuppressive agents), haemolysis (e.g. chloramine toxicity, drug-induced, oxidative) and hypersplenism.46,76 In the case of haemolysis, Gallucci et al. studied nine haemodialysis patients with ESA hyporesponsiveness, who demonstrated increased oxidative damage of the erythrocyte membrane resulting in heightened haemolysis.106 Such increased red blood cell destruction may be ameliorated by antioxidant therapies, including vitamin E.67

Patients with haemoglobinopathies, such as sickle cell disease and α- and β-thalassaemia, also tend to respond poorly to ESA. In haemoglobin SS and SC diseases, ESA treatment results in release of reticulocytes containing predominantly haemoglobin S, with little if any increment in the more stable haemoglobin F (which protects against sickling).107 Such ESA hyporesponsiveness may improve following renal transplantation.108 In patients with β-thalassaemia minor, the amount of anomalous haemoglobin chain (HbA2) directly correlates with ESA dose, strongly indicating the magnitude of resistance to ESA.109


A suggested practical approach to a patient with ESA hyporesponsiveness is depicted in Figure 1. Requesting a reticulocyte count is often a very useful initial step for identifying patients with blood loss or haemolysis, in which the reticulocyte count is frequently elevated. Such patients should be investigated with upper and lower gastrointestinal endoscopies or haemolytic screens, as appropriate. Blood loss should always be suspected in patients who require increasing doses of ESA to maintain a stable haemoglobin level, in patients whose haemoglobin levels are falling, and in patients with failure to augment iron stores in the face of repetitive i.v. iron loading. Optimal approaches for evaluating patients with obscure gastrointestinal bleeding (supported by positive faecal occult blood testing) are evolving, and recommendations have been issued in an official guideline from the American Gastroenterological Association and the American Society for Gastrointestinal Endoscopy.110 It is initially recommended that a repeat upper endoscopy (with or without push enteroscopy) is a reasonable first step followed by a repeat colonoscopy (with intubation of the terminal ileum) to identify potentially overlooked lesions (e.g. angiodysplasiae). If no cause is identified, it may be reasonable to proceed to wireless capsule endoscopy, which is highly sensitive for detecting small bowel lesions (e.g. angiodysplasia, tumours, ulcers), but has the disadvantages of not permitting tissue sampling or therapeutic intervention. These disadvantages may be overcome by performing complementary double-balloon enteroscopy.

Figure 1.

Suggested approach to patients with erythropoiesis-stimulating agents (ESA) hyporesponsiveness. ACEi, angiotensin-converting enzyme inhibitors; CRP, C-reactive protein; HRC, hypochromic red cells; LDH, lactate dehydrogenase; SC, subcutaneous; TIBC, total iron-binding capacity; URR, urea reduction ratio.

In patients with low reticulocyte counts (<40 × 109/L), compliance should be evaluated in patients self-administering an ESA. If non-compliance is effectively excluded, the two commonest causes of ESA hyporesponsiveness in this setting are iron deficiency and inflammation, which can be readily detected by ordering iron studies and serum CRP levels (normal sensitivity), respectively. Patients with absolute or functional iron deficiency should be treated with i.v. iron supplementation (provided serum ferritin levels are below 500 ng/mL). In the case of inflammation-associated anaemia, the underlying cause of inflammation should be specifically treated if able to be identified (e.g. antibiotics for infection, removal of failed renal allografts, etc.). If the TSAT is low and the ferritin level is high, increasing the ESA dosage is often useful and further i.v. iron supplementation may be beneficial, although the long-term safety of this practice is uncertain. In the absence of an identifiable cause of inflammation/infection, other potential therapeutic manoeuvres that may be considered include the use of biocompatible membranes, ultrapure dialysate, ascorbic acid therapy, vitamin E supplementation, statins or oxpentifylline administration. If iron deficiency and inflammation are excluded, investigation and management of possible underdialysis, haematinic factor deficiencies, hyperparathyroidism and aluminium overload should be undertaken. Haemoglobin electrophoresis should also be considered in the appropriate clinical setting (e.g. α-thalassaemia in Asian patients, β-thalassaemia in patients of Mediterranean origin). If a patient is receiving ACEi and/or ARB therapy and there is no other clearly definable cause for their ESA hyporesponsiveneness, it seems prudent to trial cessation of these agents if the anticipated benefits exceed the anticipated risks. Ongoing ESA hyporesponsiveness warrants a bone marrow examination to evaluate iron stores and exclude bone marrow disorders (including myelodysplasia, myelofibrosis and PRCA). If a cause of ESA hyporesponsiveness is still not identified, it may be reasonable to consider increasing the ESA dose further, although this exercise is seldom fruitful if the ESA dose already exceeds 300 IU/kg per week EPO or 1.5 μg/kg per week DPO.


Erythropoiesis-stimulating agents hyporesponsiveness occurs in approximately 5–10% of patients receiving ESA and represents an important diagnostic and management challenge. Absolute or functional iron deficiency represents the most important cause of this condition and it is widely accepted that maintaining adequate iron stores, ideally by administering iron parenterally, reduces the requirements for, and enhances the efficacy of ESA. Infection and other conditions leading to inflammatory states represent the second commonest cause of ESA hyporesponsiveness and often respond to specific treatment as well as anti-inflammatory therapies, such as ascorbic acid therapy, vitamin E supplementation, statins or oxpentifylline administration. Numerous other factors may also influence responsiveness to ESA, including dialysis adequacy, haematinic factor levels (e.g. vitamin B12, folate, carnitine), aluminium overload, hyperparathyroidism and possibly ACEi/ARB administration. Enhancing ESA responsiveness is critically important to minimizing the high cost of ESA therapy as well as improving clinical outcomes.