The ghost of parvovirus past: Idiopathic pure red cell aplasia responding to IVIG following resolved perinatal parvovirus B19 infection

To the Editor: Pure red cell aplasia (PRCA) is characterized by anemia in the setting of reticulocytopenia and markedly decreased or absent erythroid precursors in the bone marrow. Congenital PRCA is most commonly associated with Diamond–Blackfan anemia (DBA) syndrome secondary to germline ribosomal protein (RP), GATA1, or TSR2 mutations, though mutations in ADA2 and EPO have also been implicated.1-5 In comparison, acquired PRCA can be classified as either primary (idiopathic) or secondary to drugs, autoimmunity, lymphoproliferative disorders (chronic lymphocytic leukemia, large granular lymphocytic leukemia), solid tumors (thymoma), anti-erythropoietin (EPO) antibodies, or infections.6-8 The most notable viral-induced etiology of PRCA is human parvovirus B19 (B19), a highly contagious single-stranded DNA virus that is toxic to erythroid precursors and classically associated with PRCA in immunocompromised hosts.8, 9 B19 can cause transient aplastic crisis in individuals with chronic hemolysis and nonimmune hydrops fetalis when maternally transmitted to a fetus.9-11 We report an unusual case of a 5-year-old female with no syndromic features and unremarkable family history who developed recurrent transfusion-dependent anemia. The patient was born prematurely at 29 weeks gestation, presenting with nonimmune hydrops fetalis and severe anemia (hemoglobin 6.9 g/dL), secondary to maternal transmission of B19 (positive maternal B19 qualitative polymerase chain reaction [PCR] and serology [IgM 3.9 intravenously (IV), IgG 5.2 IV]). During the first 12 weeks of life, she was treated with exchange and packed red blood cell (PRBC) transfusions only. At discharge, her hemoglobin was 14.2 g/dL with a reticulocyte percentage of 1.4%, 5 days after last PRBC transfusion. Two months later, she presented with profound anemia in the setting of continued B19 infection (hemoglobin 3.3 g/dL, absolute reticulocyte count [ARC] 0.008 × 106/mm3 [0.9%], quantitative B19 PCR >100 million DNA copies/mL, IgM 5.15 IV, IgG 0.34 IV). She received intravenous immunoglobulin (IVIG) 2 g/kg in addition to four PRBC transfusions, and was discharged with a hemoglobin of 13.1 g/dL and ARC of 0.026 × 106/mm3 (0.6%). During follow-up evaluations, quantitative B19 PCR gradually decreased over 9 months until it was undetectable, coinciding with seroconversion to negative B19 IgM and positive B19 IgG. She remained asymptomatic with normal growth and development for greater than 4 years, when without evident triggers she developed chronic anemia requiring PRBC transfusions every 3−4 weeks (hemoglobin nadir 6.5–7.5 g/dL, ARC 0.010–0.015 × 106/mm3 [∼0.2%–0.5%]), except for an 8-week period without transfusion after a single dose of IVIG 1 g/kg (Figure 1). Bone marrow evaluations during transfusion-dependence revealed loss of erythroid precursors consistent with PRCA (Figure 2A,B) without dysplasia, increased blasts, or giant proerythroblasts characteristic of B19 infection.10 Complete B19 eradication was further supported by negative PCR on both bone marrow and peripheral blood and repeat serology consistent with past infection (IgM negative, IgG positive), therefore IVIG therapy was not continued. Whole genome sequencing-based VirusScan12 did not detect viral sequences from B19, CMV, EBV, HHV-6, HHV-7, HPV, HBV, or TTV. EPO was appropriately elevated at 1433 mU/mL. There was no evidence of hemolysis (total bilirubin 0.7 mg/dL, lactate dehydrogenase [LDH] 230 U/L, haptoglobin 244 mg/dL), autoimmune disease (negative DAT, ANA, ALPS panel, and complement testing), or immunodeficiency (normal lymphocyte subsets and immunoglobulins). Genetic testing for congenital PRCA (RP genes associated with DBA as well as ADA2 and EPO)5 and red cell enzymopathies was negative. Chromosome microarray was normal and whole genome sequencing (performed under the INSIGHT-HD protocol, ClinicalTrials.gov: NCT02720679) failed to identify any pathogenic variants for hematologic disease or immunodeficiency, supporting a diagnosis of acquired idiopathic PRCA. Due to continued transfusion dependence, a trial of prednisolone (2 mg/kg/day) was given for 5 weeks, as approximately 20%–30% of patients with DBA syndrome lack molecular diagnosis.1, 13, 14 Due to steroid nonresponse and previous evidence of decreased transfusion requirement status post IVIG, we proceeded with a structured course of IVIG (at 0.5 g/kg/day × 4 days). On Day 5, she was noted to have a 10-fold increase in ARC (0.011 × 106/mm3 [0.3%] to 0.112 × 106/mm3 [3.5%]), followed by hemoglobin normalization within 1 week. Bone marrow evaluation while receiving IVIG revealed trilineage hematopoiesis with normal erythroid precursors (Figure 2C,D). IVIG was continued at 0.2 g/kg/dose every 2−3 weeks, and the patient remained transfusion-independent for a period of 12 months. Discontinuing IVIG resulted in anemia relapse within 4 weeks, with hemoglobin again normalizing when IVIG was resumed (loading 1 g/kg followed by 0.2 g/kg maintenance every 4 weeks). Due to difficulties with intravenous access and the burden of IVIG infusions, we sought oral therapy targeting presumed antibody-mediated dysimmunity. Given its cytostatic effect on lymphocytes and favorable side effect profile, we trialed mycophenolate mofetil (MMF), which suppresses T-/B-lymphocyte proliferation and antibody production through its effects on purine synthesis.15 However, MMF monotherapy (500 mg/m2/dose twice daily) over 6 months failed to sustain a hemoglobin higher than 9 g/dL, our target hemoglobin for children with DBA syndrome.5 IVIG infusions were successfully reinstated at 0.2 g/kg at gradually increasing intervals of 4−16 weeks, for a total of 10 doses over 2 years before discontinuation. At last follow-up, the patient is 10 years old and remains transfusion-independent with a hemoglobin persistently higher than 13.0 g/dL (reticulocyte 1.5%–2.0%) off all therapy for more than 21 months (Figure 1). This case highlights the potential for late relapses of PRCA even after resolution of the initial trigger, such as B19 infection. While congenital PRCA (DBA) is treated with corticosteroids, chronic PRBC transfusions, or hematopoietic stem cell transplantation (HSCT), treatment for idiopathic PRCA focuses on T-cell-targeted immunomodulation, including corticosteroids, MMF, cyclosporine, cyclophosphamide, anti-thymocyte globulin, 6-mercaptopurine, monoclonal antibodies (anti-CD20, anti-CD52), splenectomy, or plasmapheresis.6-8, 15-17 Our case suggests IVIG may also be an effective treatment option for idiopathic PRCA refractory to standard therapies, even in the absence of active B19 infection or immunodeficiency. We hypothesize that the broad immunomodulatory effects of IVIG, including inhibition of Fc-mediated destruction of antibody-coated cells and cytokine neutralization,18 contributed to the restoration of normal erythropoiesis in our patient, circumventing the undesirable effects of intensive immunosuppression or HSCT. In conclusion, our case suggests that idiopathic PRCA presenting in the absence of active B19 infection, immunodeficiency, lymphoproliferative disorders, or positive molecular diagnostics may be the result of an autoimmune attack specific to early erythroid precursors. Although the efficacy of IVIG in this setting has not been widely described, it should be considered as an alternative, individualized therapeutic approach for patients with PRCA unresponsive to conventional treatments. IVIG serves as both a diagnostic tool to identify antibody-mediated autoimmune processes and a potential long-term remission-inducing therapy, avoiding the need for chronic transfusions or HSCT. However, further research is needed to better understand the underlying pathophysiology of idiopathic PRCA, establish the broader applicability and long-term efficacy of IVIG treatment, and identify patients who may benefit most from this approach. We thank Melvanique Hale, Jessica Uhrich, Kelsey Ray, and other members of the St. Jude Bone Marrow Failure Program for their assistance in clinical management. Grant support: NIH P30CA021765, Edward P. Evans Foundation, and Vera and Joseph Dresner Foundation (MWW). The authors declare no conflicts of interest.