Dramatic response to intravenous immunoglobulin in erythroblastic synartesis.

A 38-year-old man with no previous medical history was admitted to the hospital with an altered general condition. Biological tests revealed non-regenerative anaemia (haemoglobin [Hb] 5.3 g/dL, mean corpuscular volume 83.3 fL, reticulocytes 8.8 × 109/L). Ferritin, bilirubin, haptoglobin, folate, vitamin B12, protein electrophoresis, platelets, leucocytes, thyroid hormones and blood lead levels were within normal limits. Lactate dehydrogenase (LDH) were elevated at 404 iu/L (normal range: 135–246), and the direct anti-globulin test was negative. The blood film showed anisocytosis and anisochromia, the presence of a few elliptocytes, rare dacryocytes and a few agglutinated red blood cells. Paroxysmal nocturnal haemoglobinuria was excluded by flow cytometry (FLAER). Search for NUP98, MLL (KMT2A) and ETV6 rearrangement were negative. Bone marrow smears were rich, consisting mainly of numerous erythroblasts, mostly grouped in clusters (Figure 1), suggesting the diagnosis of erythroblastic synartesis. No other abnormalities were observed. The karyotype and next generation sequencing (NGS) revealed no abnormalities. A bone marrow biopsy showed erythroid hyperplasia. A minor B-cell clone was suggested by B-cell lymphocytosis (640/mm3) with 80% kappa light chain positivity. Serum-free light chain, serum immunofixation and Bence Jones protein revealed no abnormalities. A body computed tomography scan showed discrete splenomegaly. 18-Fluorodeoxyglucose positron emission tomography-computed tomography (18FDG PET-CT) was normal. After an initial course of prednisone 1 mg/kg/day for 1 month associated with high-dose erythropoietin (EPO) 300 μg per week for 6 months, the patient remained dependent on red blood cell (RBC) transfusions. A treatment with rituximab (1 g of D1 and D15) was started. At 4 month after the first rituximab infusion, the patient remained dependent on RBCs transfusions. He received a total of 44 RBCs transfusion units. A complete clinical, biological and imagery workup at this time was normal except for severe non-regenerative anaemia. Eventually, intravenous immunoglobulin (IVIg, 2 g/kg/month) were administered to the patient, which led to a rapid and complete response with normalization of haemoglobin (14 g/dL), enabling red cell transfusions to be stopped (Hb >10 g/dL at 1 month and normalized at 2 months). The patient remains in complete response after 12 months of treatment, and courses of treatment could be spaced every 6 weeks after 6 months of treatment. At 12 months, IVIg was discontinued, and the patient remained with a normal blood cell count at 8 months of follow-up. Erythroblast synartesis (ES) was first described in 1973, based on an observation of septate junctions acquired by the erythroblast in a case of refractory anaemia.1 ES is a rare and probably often unrecognized cause of severe anaemia with acquired dyserythropoiesis and is frequently associated with monoclonal gammapathy and/or clonal B-cell disease. Ineffective erythropoiesis appears to be linked to erythroblastic hyperplasia and the presence of pseudo-binuclear erythroid cells, which are erythroblasts linked by interdigitated cellular processes. IgG-activated membrane antigens (CD36, integrin VLA4 and 5, E cadherin and Emp) initiate interdigitated cytoplasmic processes between erythroblasts (zipper structure), leading to the formation of erythroblastic islets images as shown in Figure 2. As a result, mature RBCs are not released and are destroyed intramedullary. ES can be mistaken for red cell dysplasia, but the presence of erythroblastic clusters without macrophages should help correct the diagnosis. This case report illustrates the difficulty to make the diagnosis of this rare disease. A few cases are described in the literature. In a recent case series of three patients with the review of nine previously reported cases, two of the three patients were males with median age of 65, median haemoglobin level was 6.7 g/dL and six patients had associated B lymphopathy: three with chronic lymphocytic leukaemia, one with low-grade B-cell lymphoma and two with small lymphocytic lymphoma. Eight had monoclonal gammapathy of undetermined significance, of which six were IgG kappa.2 The present case is not associated with a clonal lymphoproliferative disorder or a monoclonal gammapathy. Patients with ES reported were treated with different therapies, and the response to treatment was variable. In three cases, corticosteroids used alone were associated with a response, but all relapsed. Ciclosporin or rituximab were reported with complete response in two separate cases. Other patients were treated by combined therapies, including either chlorambucil, cyclophosphamide, rituximab or ibrutinib, depending on the underlying disease. The efficacy of IVIg in this pathology has not been reported yet. Only one case reported the use of IVIg in ES without an associated monoclonal disorder. In this case, IVIg and prednisone were ineffective, but eventually the patient responded to anti-CD20.3 Here, it was associated with a dramatic and long-lasting complete biological response, eventually allowing treatment discontinuation. IVIg exhibits a large number of immune modulatory properties mediated either by the Fc or the Fab portion and is used for these properties in several autoimmune disorders, including immune thrombocytopenia.4 Because ES pathophysiology is largely unknown, it is difficult to speculate on the mechanism of action of IVIg. However, the rapid recovery after the first infusion (reappraisal of reticulocytes after days 18) suggests a direct mechanism for resolving the overexpression of erythroblast adhesion molecules induced by immune dysregulation, maybe through IgG neutralization and/or Fc receptor saturation. Given this observation and the safe profile and tolerance of IVIg,4 we suggest that IVIg could be proposed to patient with ES. Clara Fontenaille: Writing—original draft. Benjamin De Sainte Marie: Writing—review and editing. Marie Loosveld: Writing—review and editing. Mikael Ebbo: Writing—review and editing. Nicolas Schleinitz: Conceptualization, supervision, writing—review and editing. All authors contributed to patient care and writing of the report. The authors have no conflicts of interest to disclose. The authors confirm that the data supporting the findings of this study are available within the article. Raw data that support the findings of this study are available from the corresponding author.