HLA-B*46:01:01:01 and HLA-DRB1*09:01:02:01 are associated with anti-rHuEPO-induced pure red cell aplasia

Abstract Treatment of anemia in patients with chronic kidney disease (CKD) with recombinant human erythropoietin (rHuEPO) can be disrupted by a severe complication, anti-rHuEPO-induced pure red cell aplasia (PRCA). Specific HLA genotypes may have played a role in the high incidence of PRCA in Thai patients (1.7/1,000 patient years vs. 0.03/10,000 patient years in Caucasians). We conducted a case–control study in 157 CKD patients with anti-rHuEPO-induced PRCA and 56 controls. The HLA typing was determined by sequencing using a highly accurate multiplex single-molecule, real-time, long-read sequencing platform. Four analytical models were deployed: Model 1 (additive: accounts for the number of alleles), Model 2 (dominant: accounts for only the presence or absence of alleles), Model 3 (adjusted additive with rHuEPO types) and Model 4 (adjusted dominant with rHuEPO types). HLA-B*46:01:01:01 and DRB1*09:01:02:01 were found to be independent risk markers for anti-rHuEPO-induced PRCA in all models [OR (95%CI), p-values for B*46:01:01:01: 4.58 (1.55–13.51), 0.006; 4.63 (1.56–13.75), 0.006; 5.72 (1.67–19.67), 0.006; and 5.81 (1.68–20.09), 0.005; for DRB1*09:01:02:01: 3.99 (1.28–12.49), 0.017, 4.50 (1.32–15.40), 0.016, 3.42 (1.09–10.74), 0.035, and 3.75 (1.08–13.07), 0.038, in Models 1–4, respectively. HLA-B*46:01:01:01 and DRB1*09:01:02:01 are susceptible alleles for anti-rHuEPO-induced PRCA. These findings support the role of HLA genotyping in helping to monitor patients receiving rHuEPO treatment.