Bone marrow necrosis in childhood: The hunt for blasts.

A 6-year-old child born to non-consanguineous parents, without notable family health history, presented to the emergency department with febrile diffuse bone pain. The hemogram showed isolated thrombocytopenia (48 × 109/L), a mild leukocytosis (leukocytes at 12 × 109/L) with polymorphonuclear neutrophils at 7.08 × 109/L. The blood smear disclosed 8% lymphoid blasts (Figure 1A). Peripheral blood flow cytometry identified a blastic population of common B cell precursor acute lymphoblastic leukemia (CD19+, CD22+, CD79a+, CD10+, cμ−, sIg−) partially coexpressing CD13. Two successive bone marrow aspirations were performed on the anterior superior iliac spine and showed bone marrow necrosis without identification of any hematopoietic cells nor blast cells (Figure 1B,C). The standard karyotype and fluorescence in situ hybridization were unsuccessful because of bone marrow necrosis. The same tests carried out on peripheral blood samples were unsuccessful. Surprisingly, next-generation sequencing revealed a BCR::ABL1 fusion, joining intron 1 of BCR and intron 1 of ABL1 associated with deletions of the IKZF1 and PAX5 genes. The patient received intensive chemotherapy with prednisone, vincristine, PEG-asparaginase, daunoribucine in association with imatinib. The induction phase was complicated by severe colitis, septic shock due to E. coli, and acute respiratory distress syndrome due to cytomegalovirus and aspergillus pneumonia. Neutrophil recovery occurred under granulocyte colony-stimulating factor (G-CSF) at D44 after the start of induction therapy. Bone marrow aspiration was performed at day 50 on the same area and showed necrosis. Immunoglobulin/TCR gene rearrangement allele-specific real-time quantitative PCR (Ig/TCR) performed at D50 detected the presence of measurable residual disease (MRD) greater than 10−2. Finally, a bone marrow aspiration performed on the posterior superior iliac spine allowed to confirm the cytological remission within a regenerative bone marrow (Figure 1D) at D53. Ig/TCR MRD a was detectable and quantified at <10−4 on this specimen. Histological analysis of a contemporary (D53) bone marrow biopsy of the posterior superior iliac spine revealed a necrotic population occupying 40% of the sample (Figure 2). Immunohistochemistry showed that the necrotic population was positive for CD34, CD10 and terminal deoxynucleotidyl transferase (TdT). The observation of bone marrow necrosis is a rare phenomenon,1 especially in pediatrics.2 In the pediatric population, cancers are the leading cause of bone marrow necrosis (acute leukemia, metastatic neuroblastoma).2 In adults, hematological malignancies seem to be frequent causes of bone marrow necrosis, and are associated with a poor prognosis.1 Non-malignant etiologies include severe infections, autoimmune diseases, disseminated intravascular coagulation, thrombotic disorders, and G-CSF exposure.3 In ALL at diagnosis, local ischemic processes or the pressure of blast cells in the inextensive bone marrow cavity could explain the necrosis phenomenon.3 In the case presented here, the discrepancy between the cytological complete remission post induction and the presence of lymphoblastic necrotic cells on the bone marrow biopsy raises the question of the relevance of the Ig/TCR MRD, which is unable to discriminate between residual blasts and necrotic cells. Furthermore, there appears to be spatial heterogeneity of repartition in bone marrow necrosis in our patient. Thus, the presence of bone marrow necrosis, particularly in children, should raise suspicion for the diagnosis of acute lymphoblastic leukemia and warrant further investigation. Immunochemistry may be helpful in characterizing the necrotic cells.1 A careful examination of the blood smear, possibly supplemented by flow cytometry, is also essential. Baptiste Le Calvez, Marion Eveillard, Margaux Camuset and Audrey Grain wrote the manuscript. Marion Eveillard performed experiments. Margaux Camuset and Audrey Grain treated the patient. All the authors critically reviewed the manuscript. The authors are grateful to Pr. H. Cavé and Dr. A. Caye for performing some of the molecular investigations and to Dr. J. Lamer for the histological analysis. The research work of Baptiste Le Calvez is financed by the association Etoile de Martin/SFCE. The authors have no conflict of interest to disclose. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.