Multiple Mutations of CEBPA Contribute to Leukemogenesis: The Donor Origin of Leukemia Relapse After Allogeneic Hematopoietic Cell Transplantation.

Leukemia relapse arising in cells of donor origin in the transplant recipient, called donor cell leukemia (DCL), is a rare disease entity after allogeneic hematopoietic stem cell transplantation (allo-HSCT). The precise etiological mechanisms of DCL are unravelled and almost all the reported cases do not suggest a common mechanism. Careful analyses of the mechanisms with respect to the oncogenic transformation of donor-derived cells might provide a valuable insight into understanding of leukemogenesis.We aimed to assess whether those genetic mutations implicated in the development of common forms of AML contribute to the “leukemization” of donor cells in DCL.

(1) A 36-year-old male was diagnosed with AML-M4. Cytogenetic evaluation demonstrated an abnormal clone 46 XY, del (9) (q11 q34) in 10/10 cells. The patient underwent `allogeneic peripheral blood stem cell transplantation (allo-PBSCT) from a HLA-identical sister. Short tandem repeats (STR) analyses on day +28 showed complete donor chimera. Thirteen months after SCT, bone marrow aspiration revealed leukemia relapse. STR analyses showed still absolute donor chimera. The karyotype of bone marrow cells of the patient showed a new clonal chromosome abnormality: 45,XX, der(15;22) (q10;q10) in 10/10 cells, which was completely identical to the karyotype of the donor. Molecular evaluation suggested the patient developed DCL from a HLA-identical sibling. The examination of the donor's bone marrow showed normal and no malignant clone was detected by flow cytometry and fluorescence in-situ hybridization analyses. The donor remains healthy during a 25-month follow-up. (2) A series of archival stained bone marrow slides of the patient, including at the times of diagnosis, CR after one course of induction chemotherapy, lasting CR, before SCT, 1 month, 9 months and 12 months after SCT, samples of mononuclear-cell-enriched bone marrow at the times of relapse and CR after relapse, and buccal mucosal swab specimen during remission were available. (3) Buccal mucosal swab specimen and samples of mononuclear-cell-enriched peripheral blood and bone marrow were available from the donor. (4) Genomic DNA was extracted and analyzed for mutations in fms-related tyrosine kinase 3 gene (FLT3), neuroblastoma RAS viral oncogene homolog gene (NRAS), the CCAAT enhancer-binding proteinα gene (CEBPA), myeloid-lymphoid or mixed-lineage leukemia gene (MLL), and nucleophosmin gene (NPM1).

(1) DNA obtained from the patient during diagnosis was found to exhibit three different mutations in CEBPA, the gene encoding the crucial granulocytic differentiation factor C/EBPα. The mutation1 was the duplication of a cytosine residue at nucleotide 247 (247dupC) which resulted in the corresponding protein terminating prematurely at codon106 (Gln83fsX106). The mutation2 was a 6-bp duplication comprised nucleotides 584 to 589 (584_—589dup) which resulted in an internal tandem duplication of amino acids 195 to 196 (His195_—Pro196dup) in the protein. The mutation3 was a 3-bp duplication comprised nucleotides 914 to 916 (914_—916dup) which resulted in the duplication of amino acids 305 (Gln305dup) in the protein. The 584_—589dup mutation was remain found in DNA from the patient during lasting remission before SCT and from his buccal swab specimen, indicating this mutation should be germ-line mutation. In contrast, other two mutations were not found in the patient during remission and buccal swab specimen, indicating these two mutations should be somatically acquired. (2) The 584_—589dup germ-line mutation was also found in DNA from peripheral blood cells, bone marrow cells and buccal swab specimen from the donor. (3) Donor-derived cells in the patient at the times of 1 month, 9 months and 12 months after SCT only exhibited the single germ-line mutation. When the patient relapsed with DCL, donor-derived leukemic cells were found to develop two somatic mutations in CEBPA in the patient microenvironment, which were completely identical to those observed in the patient with de novo leukemia. No other mutations were identified in FLT3, NRAS, MLL and NPM1 genes.

Our findings suggest that multiple mutations of CEPBA impairing C/EBPα function may be sufficient to induce AML and a single mutation of CEBPA is not leukemogenic, but is the first step and predisposed to development of other mutations, ultimately inducing AML.

No relevant conflicts of interest to declare.