Background

Donor leukemia is a rare condition and only anecdotal reports described its development [1, 2]. There are two types of donor leukemia, literally the donor itself develo** leukemia or donor cells in the recipient develo** leukemia (Donor Cell Leukemia, DCL), which is more commonly reported. Two mechanisms have been proposed for DCL development: occult transfer of malignant cells from donors during hematopoietic stem cell transplantation (HSCT) or leukemic transformation of healthy donor cells in recipients [3,4,5,6]. The latter one is unpredictable but it takes up the majority of DCL. Donor develo** leukemia with recipient intact is even a rarer condition and only few reports described the development of acute myeloid leukemia (AML), not acute lymphoblastic leukemia (ALL), in peripheral blood stem cell (PBSC) donors [7, 8]. G-CSF administration was suspected for leukemogenesis but there have been no valid evidence to support the hypothesis. Meanwhile, there have been no reports describing the development of ALL in hematopoietic stem cell donors, not recipients.

Here we reported a case with only the donor develo** ALL 8 yrs after donating hematopoietic stem cells but not the recipient. Whole exome sequencing (WES) showed leukemic genetic lesions (SF3B1 and BRAF mutation) only appeared in the donor’s hematopoietic cells. A special type of HSCT using the recipient’s PBSCs was done. To our knowledge, this is the first case report of donor develo** ALL and an unusual type of HSCT. We hope it could provide more evidence for donor leukemia formation and the potential role of SF3B1 and BRAF mutation in ALL.

Case presentation

In May 2017, a 51-year old female presented to our hospital with over 1 week of fatigue. She has medical history of donating both bone marrow and PBSCs 8 yrs ago for her brother with severe aplastic anemia (SAA). G-CSF was used to mobilize hematopoietic stem cells for PBSC collection. Complete blood count (CBC) showed lymphocytosis (Lymphocyte 10*10^9/L), moderate anemia (Hb 61 g/L) and thrombocytopenia (PLT 20*10^9/L). Peripheral blood smear demonstrated 30% of blast cells. Bone marrow biopsy showed lymphoblastic leukemia involving a markedly hypercellular marrow (Fig. 1). Flow cytometry showed the blasts expressed CD10, CD19, CD22, CD38, HLA-DR, but not CD13, CD33, CD117, or cytoplasmic MPO. Chromosomal analysis showed a normal female karyotype. The diagnosis of B-ALL was made. VDP regimen (Vindesine, Daunorubicin, Prednisone) was started after her diagnosis and bone marrow aspirate 14 days later revealed complete remission (CR). Then 2 cycles of VDLP regimen (Vindesine, Daunorubicin, Prednisone, Pegaspargase) and 2 cycles of MA (Mitoxantrone, Cytosine arabinoside) + Pegaspargase (PEG-Asp) regimen were administered. During this period, the patient remained in remission and cerebrospinal fluid (CSF) remained clear after intermittent four time lumbar punctures. Minimal residue disease (MRD) monitoring by flow cytometry after each cycle did not detect blast cells with abnormal phenotype. Considering she has no other HLA-identical siblings or unrelated donors from China Bone Marrow Bank at that time and her brother, who was perfectly stable, a special type of auto-HSCT using her brother’s PBSCs was performed in December 2017. Before the transplantation, whole exome sequencing (WES) was done for both the donor and recipient to rule out occult genetic abnormalities in the stem cells. Results showed that only the patient, not her brother, has genetic mutations including SF3B1 and BRAF mutation associated with hematological malignancies in her hematopoietic cells, not oral mucosal cells (Table 1) [9,10,11]. Myeloablative conditioning regimen (Melphalan 140 mg/m2 d-3/d-2/d-1, Cytarabine 1 g/m2 d-3/d-2, Cyclophosphamide 60 mg/Kg d-3/d-2) was used. A total of 2.34*10^6/kg CD34+ cells, 13.0*10^8/kg MNCs from peripheral blood were collected and transfused into the patient. Neutrophil engraftment occurred at day + 11 and platelet engraftment occurred at day + 14. VP regimen (Vindesine, Prednisone) and MTX + 6-MP regimen were used for maintenance therapy. Bone marrow aspirate and flow cytometry showed that the patient remained in complete remission 3 months after transplantation. In May 2018, the patient’s CBC showed leukocytosis and bone marrow aspirate indicated disease relapsed with 53.33% blast cells. Flow cytometry showed the blasts expressed CD10, CD19, HLA-DR and partially expressed CD11b. Genetic testing using polymerase chain reaction (PCR) technique found that SF3B1 and BRAF exome mutation was negative. Re-induction chemotherapy with VDP regimen was administered, the patient achieved CR. However, after another 2 cycles of high-dose MTX+ PEG-Asp regimen, disease relapsed. In November 2018, CD19 CAR-T therapy followed by HLA-identical unrelated hematopoietic stem cell transplantation was applied and the patient remains in remission for 7 months till now.

Fig. 1
figure 1

Bone marrow biopsy of the patient showed a markedly hypercellular marrow implying ALL

Full size image
Table 1 Genetic mutations in the sister’s hematopoietic cells (Sample A), oral mucosal cells (Sample B) and the brother’s hematopoietic cells (Sample C) detected by WES
Full size table

Discussion and conclusion

Leukemogenesis is a complex process involving accumulation of genetic changes in self-renewing hematopoietic stem cells (HSCs) [12]. Two-hit model was proposed for the clonal evolution of leukemia [13]. Genome sequencing indicated that most mutations in the leukemia genome were random events and the acquirement of one or two initiating mutations generated the founding clone for leukemogenesis [23]. In this case, we found BRAF G466E in the patient, not her brother. It might form double hits with SF3B1 mutation, which could affect genes involved in pre-B cell differentiation, and ultimately leads to leukemogenesis in this patient. More basic research needs to be done to clarify the mechanism.

ZC3H7B and LAMC1 gene mutations were also found in the patient’s hematopoietic cells, but there have been no reports of them in hematological malignancies. Their roles in leukemogenesis still need to be uncovered.

G-CSF is used in the majority of allogeneic hematopoietic stem cell donations. Some studies showed its application was associated with increased risk of later development of MDS/AML [6, 8]. Meanwhile, some other studies failed to identify the correlation [24, 25]. It is a critical issue but it remains controversial. In our case, the donor developed ALL 8 yrs after hematopoietic stem cell mobilization using G-CSF. Her leukemogenesis and G-CSF usage were probably irrelevant since G-CSF is a regulator for granulopoiesis, not lymphopoiesis. The duration and dosage of G-CSF administration was longer and higher in the recipient brother, but till now no signs of MDS/AML have been detected in him. Therefore, high quality and more conclusive clinical data is needed to clarify G-CSF usage and leukemia formation especially in healthy donors.

The suspicion of delayed onset of DCL was also considered. According to previous reports of DCL, the average latent period is within 4 yrs of HSCT [26, 27]. In our case, the recipient brother did not develop leukemia 8 yrs after transplantation. Additionally, WES showed no leukemia-associated gene mutations in him and it further ruled out the possibility of delayed leukemia onset.

Auto-HSCT can be considered in MRD negative ALL patients without appropriate HLA-identical donors. In this case, a special type of auto-HSCT was performed. Compared with stem cells collected from the patient herself, peripheral blood stem cells from her brother were leukemia cell free and it was supposed to lower the risk of disease relapse. However, the patient relapsed 3 months after transplantation. Since SF3B1 and BRAF mutation turned negative after disease relapsing, we hypothesized that leukemia sub-clones outgrew and caused disease relapse.

To our knowledge, this is the first case report with only donor develo** ALL but not the recipient and this is also the first report of an unusual type of auto-HSCT. We hope it could provide more evidence for the double-hit model in leukemogenesis and a potential alternative for donor leukemia treatment.