Congenital Undifferentiated Pure Erythroid Leukemia

INTRODUCTION: Congenital pure erythroid leukemia (M6b) is exceedingly rare with only a few reported cases to date. Because of the extreme rarity, almost nothing is known about the pathogenesis, appropriate therapy and prognosis. Diagnosis of erythroid leukemia is usually based on the positivity for Glycophorin A, Glycophorin C or PAS staining. We report a first case of congenital pure erythroid leukemia expressing E-cadherin in the absence of Glycophorin A, Glycophorin C and PAS staining. We analyzed the cytogenetic abnormalities of this extremely rare disease.

RESULTS: The patient was the first daughter of healthy and non-consanguineous Japanese parents, born at 40 weeks of gestation by emergency cesarean section in non-reassuring fetal state after uncomplicated pregnancy. Apgar score was 8/9. Characteristic facial appearance was not recognized. At birth, she presented with marked hepatomegaly, purpura and disseminated intravascular coagulation. White blood cell (WBC) count was 63.5x10⁹/L with blastic cells with vacuoles. Although congenital leukemia was suspected, flow cytometric analyses using CD45 blast gating failed to demonstrate leukemic cells. Karyotype was 46, XX. Fluorescence in situ hybridization (FISH) for trisomy 21 and MLL split signal were negative. GATA1 mutation was not detected. WBC count has gradually decreased within 3-4 weeks with supportive care.However, liver failure, hemophagocytic lymphohistiocytosis and schistocytosis developed. Although treatment with dexamethasone and etoposide has started, multiple nodules appeared in the liver 11 weeks after birth. Liver biopsy demonstrated small round cell tumor with high N/C ratio and vacuoles infiltrating the liver. The tumor cells were immunohistochemically positive for CD43, CD71, E-cadherin, beta-catenin, Ki-67 and c-Myc and negative for CD45, CD20, CD10, PAX5, CD3, CD4, CD8, TdT, CD1a, CD34, CD56, cyMPO, c-kit, CD42b, CD61, Glycophorin A, Glycophorin C, tyrosine hydroxylase, PGP9.5, myogenin, glypican3, NKX2.2, CAM5.2 and Periodic Acid Schiff (PAS) staining etc. Flow cytometric analysis revealed CD43+ CD71+ CD36+ CD58+ cells within large CD45 negative cell population. These cells expressed almost no other hematopoietic cell markers used to screen for leukemia. These cells were indistinguishable from normal erythroblast based on surface markers only. However, flow cytometric cell sorting revealed these cells are blasts with vacuoles. Karyotype of tumor cells has changed to 50, XX, +7, +8, add(15)(q22), +19, add(19)(q13.1-13.3)×2, +21. Based on these results, she was diagnosed with pure erythroid leukemia. Low dose Cytosine Arabinoside improved her clinical symptoms. She is alive at 5 months of age.

DISCUSSION: E-cadherin is a selective marker of immature erythroblast. In our case, E-cadherin was key in erythroid lineage assignment. To our knowledge, this is the first reported case of infantile pure erythroid leukemia expressing E-cadherin in the absence of Glycophorin A, Glycophorin C and PAS staining. These results suggest that the tumor cells originated from undifferentiated erythroblast. This disease entity should be recognized. Immunohistochemical staining of c-Myc showed strong positivity. The c-Myc gene is located on chromosome 8. FISH for c-Myc split signal was negative. G-banding and FISH revealed trisomy 8. Overexpression of c-Myc may be involved in the pathogenesis of this undifferentiated pure erythroid leukemia. At birth, karyotype was 46, XX and blasts in peripheral blood decreased with supportive care only. However, we observed changes in karyotype of blasts. We assume that second hit was added during clinical course. Whole exome sequencing analysis is in progress to reveal somatic and germline mutations underlying this unrecognized disease.