Anti-erythropoietin antibody–mediated pure red cell aplasia in a patient with chronic kidney disease

Pure red cell aplasia (PRCA) is a rare disorder characterised by normocytic, normochromic anaemia, severe reticulocytopenia and selective absence of erythroid precursors in the bone marrow, while myeloid and megakaryocytic lineages remain unaffected.¹ Acquired PRCA in adults may be idiopathic or associated with autoimmune diseases, viral infections (notably parvovirus B19), lymphoproliferative disorders, thymoma or certain medications.²

In patients with chronic kidney disease (CKD), a specific and rare form of PRCA is caused by the development of neutralising anti-erythropoietin (EPO) antibodies. We report the case of a patient with progressive CKD and transfusion-dependent anaemia who was investigated for PRCA in the setting of EPO resistance and reticulocytopenia.

Case report

Mrs B is a 79-year-old woman with a background of IgA nephropathy and progressive CKD (stage G4A2) who presented with 3 days of worsening nausea, vomiting, shortness of breath on exertion and an acute kidney injury. Her serum creatinine was 320µmol/L, elevated from a baseline of 200–220µmol/L, and haemoglobin was 66g/L on admission. On examination, she was pale but hemodynamically stable, with mild chronic peripheral oedema and no focal signs of infection or bleeding.

Mrs B had a long-standing microcytic, hypochromic anaemia, with no identified cause despite extensive evaluation. Gastroscopy and colonoscopy revealed mild gastritis and polyps but no active bleeding. Haematologic investigations showed normal iron studies, B12, folate, liver function tests and thyroid function. A haemolysis screen and serum electrophoresis were negative. Reticulocyte counts were persistently low (3×10⁹/L), suggesting suppressed erythropoiesis.

Notably, Mrs B had been started on recombinant erythropoietin (rHuEPO) in the form of binocrit from Sandoz Novartis (biosimilar of epoetin alfa) a year prior (November 2023) when her haemoglobin fell below 100g/L. Since then, she had required multiple admissions for symptomatic anaemia and had been transfused seven units of red blood cells throughout 2024 prior to the current admission. Her rHuEPO had been up-titrated to a maximal dose over the course of the following year. Despite escalating binocrit from 6,000 units/week to 30,000 units/week (10,000 units three times weekly), there was no improvement in haemoglobin levels.

Mrs B received two units of red blood cells with only temporary symptomatic relief. Both haematology and renal services were involved given the presence of erythropoiesis-stimulating agent (ESA) resistance and lack of clear explanation for her anaemia. It was felt by the renal service that the degree of anaemia and lack of response to rHuEPO could not be explained by her CKD alone. Possible differentials suggested by haematology included sideroblastic anaemia or an underlying myelodysplastic syndrome (MDS), as well as PRCA due to her CKD history and Binocrit use.

Further investigations recommended by haematology included parvovirus DNA and ANA, both of which were negative. An erythropoietin assay was also conducted, with a level of 6.8mU/mL. A non-contrast computed tomography abdomen/pelvis showed no intra-abdominal pathology or malignancy to account for her anaemia. A bone marrow biopsy was subsequently performed, which showed normocellular bone marrow, containing no erythroid precursors. There was no increase in blast cell numbers or abnormal lymphoid population to suggest a lymphoproliferative disorder. The changes seen in the bone marrow were thought to be consistent with PRCA, with aetiology most likely to be secondary to anti-EPO antibodies following Binocrit therapy.

Following provisional diagnosis, Mrs B was discharged with outpatient renal follow-up. A sample for EPO antibodies was sent to Germany (MVZ Labor PD Dr. Volkmann und Kollegen SE & Co. eGbR, Germany in December), which later returned a positive result. Mrs B was treated with prednisone 40mg daily for 4 weeks but remained transfusion dependent, requiring one unit of red blood cells every 2 weeks. Given her lack of improvement, she was started on a hypoxia-inducible factor prolyl hydroxylase inhibitor (HIF-PHI) in the form of roxadustat 50mg three times weekly. Her haemoglobin has now remained stable above 90g/L.

Discussion

Anaemia in CKD is primarily driven by inadequate EPO production from impaired renal function, leading to insufficient stimulation of erythroid precursors. While renal anaemia is a common and expected complication of CKD, it typically shows at least partial responsiveness to iron supplementation and escalating ESA therapy. Failure to respond to high-dose rHuEPO, combined with profound reticulocytopenia and exclusion of other common causes, should prompt consideration of PRCA as a differential diagnosis.

EPO antibody–mediated PRCA secondary to ESA use is rare, with the worldwide incidence being just 0.02–0.03 per 10,000 patient-years.³ A study on the incidence of EPO antibody–mediated PRCA found that out of 15,333 patients undergoing ESA therapy, only five went on to develop PRCA secondary to anti-EPO antibodies.⁴ In New Zealand, there have been 11 case reports of PRCA following treatment with EPO as of 2018, according to the Centre for Adverse Reactions Monitoring.⁵

In antibody-mediated PRCA, neutralising anti-EPO antibodies are produced in response to the presence of rHuEPO. These antibodies inhibit both endogenous and exogenous EPO, resulting in profound suppression of erythropoiesis and transfusion dependence.^6,7 Bone marrow biopsy typically demonstrates selective absence of erythroid precursors with preserved myeloid and megakaryocytic lineages.¹ It is important to exclude other causes of erythroid aplasia such as parvovirus B19 infection and haematological disorders including MDS and sideroblastic anaemias.²

Management focusses on cessation of rHuEPO to stop antigenic stimulation, combined with immunosuppressive therapy to reduce antibody production. Corticosteroids are often first-line treatment, although response may be incomplete in some cases.⁸ HIF-PHIs, such as roxadustat, represent emerging therapies that may bypass antibody-mediated EPO neutralisation. Roxadustat works by inhibiting the HIF prolyl hydroxylase enzyme, stabilising hypoxia-inducible factors (HIFs), which in turn up-regulate endogenous EPO production, enhance red blood cell synthesis and improve iron metabolism, ultimately addressing the multifactorial anaemia seen in CKD.^9,10

In addition to stimulating endogenous EPO production, roxadustat activation of hypoxia-inducible factor pathways improves iron mobilisation by suppressing hepcidin, which leads to increased absorption and better utilisation of iron stores.¹¹ This leads to increased iron availability for erythropoiesis.¹¹ HIF stabilisation also enhances erythropoiesis through effects on erythroid progenitor survival, maturation and EPO receptor expression in the bone marrow, independent of circulating EPO.¹¹

HIF stabilisation, via roxadustat or hypoxia, enhances erythropoiesis via increasing the motility of erythroid progenitor cells, leading to enhanced contact with bone marrow stromal cells.¹² This inhibits apoptosis of erythroid progenitor cells, resulting in increased absolute numbers of these cells.¹²

A meta-analysis that compared the effectiveness and acceptability of all available HIF-PHIs and ESAs found that roxadustat achieved higher haemoglobin levels than rHuEPO and reduced the risk of red blood cell transfusion rate; however, it was associated with a higher discontinuation rate secondary to adverse effects.¹³ Roxadustat has also been shown to be superior to epoetin alfa in lowering low-density lipoprotein and total cholesterol levels,¹³ which may make this a better choice for CKD patients at higher risk for cardiovascular disease.

In summary, recombinant EPO–associated PRCA remains a rare but serious complication in the treatment of anaemia, particularly in patients with CKD. Its diagnosis poses a significant clinical challenge, often requiring a high index of suspicion and thorough exclusion of other causes. While cessation of rHuEPO is the cornerstone of management, therapeutic responses to immunosuppression vary, and evidence guiding optimal treatment remains limited due to the rarity of the condition. Ongoing research and case reporting will be vital to improving recognition, diagnosis and management strategies in the future.