Remarkable response to roxadustat in a case of anti-erythropoietin antibody-mediated pure red cell aplasia

Dear Editor,

Pure red cell aplasia (PRCA) is a rare but serious complication occurring in patients with chronic kidney disease (CKD) who are treated with recombinant human erythropoietin (rHu-EPO). Patients typically develop an abrupt onset of severe transfusion-dependant anemia. Merely withdrawing rHu-EPO is not sufficient to reverse the condition, and immunosuppressive treatment with corticosteroids, cyclophosphamide, or ciclosporin only partly corrects anemia [1]. Roxadustat, an oral inhibitor of hypoxia-inducible factor prolyl hydroxylase, has showed efficacy in correcting anemia in patients with CKD who are receiving or not receiving dialysis [2, 3]. The present report is the first to describe a case of PRCA related to rHu-EPO whose hemoglobin reversed dramatically after switching to roxadustat treatment.

A 49-year-old man developed chronic renal failure due to obstructive nephropathy. He did not have other comorbidities. In November 2018, his serum creatinine level reached 631 μmol/L and hemoglobin (Hb) concentration decreased to 64 g/L. He received peritoneal dialysis and was administered 6000 units of rHu-EPO subcutaneously twice a week. Anemia improved substantially, his Hb level increased to about 100 g/L for 5 months. In June 2019, his Hb suddenly fell to 46 g/L without evidence of bleeding, infection, and hemolysis. In response, rHu-EPO was increased to 6000 units 3 times a week. In the following 5 months, he was admitted to hospital monthly for transfusion of red blood cells, his anemia improved temporarily after transfusion, and the levels of Hb at each admission fluctuated between 32 and 49 g/L. In late October, further laboratory examinations were performed. He had Hb concentration of 49 g/L, reticulocyte count of 3.4 × 10⁹/L. The level of serum EPO was below detectable limits (normal value 4.3–29.0 IU/L). Bone marrow aspirate showed severe hypoplasia of the erythroid line with hyperplasia of the granulocytic and megakaryocytic lines, normal ratio of blast cells, and no morphologic dysplasia. Assays for anti-DNA Ab, anti-nuclear Ab, hepatitis B antigen, and parvovirus B19 DNA were not detected. Computed tomography (CT) of the chest and abdomen found nothing abnormal. Anti-EPO antibody was checked by ELISA and turned out to be positive. Collectively, these findings pointed to a diagnosis of anti-EPO antibody-mediated PRCA. On November 7, 2019, rHu-EPO administrations were interrupted and he was prescribed oral roxadustat (100 mg thrice weekly). On January 21–24, 2020, he received transfusions of 5.5 units of RBCs; the Hb levels rose from 39 to 62 g/L. After then, no further transfusions were required. On February 20, 2020, his Hb level increased to 108 g/L with a reticulocyte count of 94.3 × 10⁹/L. EPO level in the serum rose to 48.9 IU/L. Bone marrow aspiration smear was hypercellular and demonstrated 19% of erythroid cells. In the following 6 months, blood routine tests showed the Hb levels increased steadily to normal value and fluctuated between 121g/L and 138 g/L (Fig. 1).

PRCA is an isolated disorder of erythropoiesis characterized by normocytic and normochromic anemia, scarcity of reticulocyte, and absence of erythroid precursors in bone marrow. Besides above features, patients developing PRCA due to anti-EPO antibodies have typically been on rHu-EPO therapy for 6 to 18 months, and anti-EPO antibodies can be demonstrated in their serum samples [4]. Regarding our patient, the anemia worsened with evidence of erythropoietic failure after administration of rHu-EPO for 7 months, and increasing the dosing frequency proved ineffective, which is in line with features of anti-EPO antibody-induced PRCA. Tacey et al. reported that EPO in serum samples with measurable anti-EPO antibody levels will be bound to the antibody, resulting in interference and underestimation of EPO by the ELISA assay [5]. Before administration of roxadustat to the patient, the serum EPO was undetectable although he was receiving treatment with rHu-EPO. After switching to roxadustat for 3.5 months, his Hb and reticulocyte count rose dramatically accompanied by high level of serum EPO, which indicated that anti-EPO antibodies decreased markedly or disappeared. Available data suggest that discontinuation of rHu-EPO alone is usually insufficient to induce recovery from anti-EPO antibody-mediated PRCA. Nine patients in a retrospective analysis had not received any specific treatment other than ceasing rHu-EPO. All nine patients did not recover from PRCA and still had detectable levels of antibodies to EPO after a median follow-up of 12 months [1]. As to our patient, we speculate that endogenous EPO upregulated by roxadustat contributed to neutralizing anti-EPO antibodies but not boost formation of antibodies to EPO. Retrospective analysis mentioned above reported all six patients who received renal transplantations recovered within 1 month and achieved normal Hb concentrations [1]. However, whether the rapid recovery is attributed to strong immunosuppression after transplantation or is due to exposure to endogenous EPO produced by the transplant is yet unclear. Observations from our patient who has not been treated with any immunosuppressant partly support the viewpoint that continuous production of endogenous EPO plays an important role in the recovery of erythropoiesis in anti-EPO antibody-induced PRCA patients after renal transplantation.

In conclusion, roxadustat was used to successfully treat a patient with anti-EPO antibody-mediated PRCA. More extensive studies are needed to reveal the exact mechanisms and answer whether there is a synergetic effect between roxadustat and immunosuppressants.