Acute Megakaryoblastic Leukemia Developing as Donor Cell Leukemia after Umbilical Cord Blood Transplantation.

A 64-year-old man with acute myeloid leukemia underwent umbilical cord blood transplantation (UCBT). After 11 months of complete remission (CR) following UCBT, the bone marrow showed 7.5% myeloblasts. CR was obtained after a single course of azacitidine monotherapy, but the myeloblasts gradually increased in the blood. We made a diagnosis of acute megakaryoblastic leukemia derived from donor cell with a fluorescence in situ hybridization (FISH) analysis of the sex chromosomes and an immunophenotypic analysis. Azacitidine was administered again and produced a therapeutic effect of stable disease. This case suggests that azacitidine may be a useful therapy for patients with acute megakaryoblastic leukemia in situations in which intensive chemotherapy and transplantation are not indicated.

Introduction

Acute megakaryoblastic leukemia has been described as a subtype of acute myelogenous leukemia (AML) (1) and was incorporated into the French-American-British (FAB) classification of AML as M7 (2). The associated bone marrow characteristics include the proliferation of abnormal megakaryoblasts identified by the presence of platelet-specific surface glycoprotein and frequently extensive myelofibrosis. Histopathologic evidence of M7 AML itself is an independent poor prognostic factor for the overall survival (3).

Azacitidine, a demethylation agent, has demonstrated a significant survival benefit in patients with high-risk myelodysplastic syndrome (MDS) relative to conventional care and has also been proved effective in older patients with newly diagnosed AML with >30% blasts (4,5). However, whether or not azacitidine is effective for acute megakaryoblastic leukemia is unclear, as it is rare in adults.

No reports are available on acute megakaryoblastic leukemia developing as donor cell leukemia (DCL) after allogeneic hematopoietic stem cell transplantation (HSCT) and treatment with hypomethylating agents (azacitidine or decitabine). We herein report a case of acute megakaryoblastic leukemia developing as DCL after umbilical cord blood transplantation (UCBT) that was treated with azacitidine.

Case Report

A 64-year-old man had pancytopenia and was referred to our hospital (Table 1). An analysis of bone marrow aspirate revealed that myeloblasts comprised 49.5% of nucleated cells with multilineage dysplasia, so we made a diagnosis of AML with myelodysplasia-related changes. He achieved hematological complete remission (CR) after induction chemotherapy consisting of idarubicin and cytarabine and underwent UCBT with a reduced-intensity conditioning regimen from an unrelated female donor in first CR 8 months after the initial diagnosis.

Table 1.

The Laboratory Data at the Onset of Original Leukemia.

Blood biochemistry			Fe	201	μg/dL
Total protein	7.2	g/dL	UIBC	184	μg/dL
Albumin	4.4	g/dL	Ferritin	144	ng/mL
BUN	12.8	mg/dL	IgG	1,197	mg/dL
Creatinine	0.94	mg/dL	IgA	400	mg/dL
AST	20	IU/L	IgM	54	mg/dL
ALT	28	IU/L	TSH	0.63	μIU/mL
LDH	175	IU/L	Free-T3	2.63	pg/mL
ALP	210	IU/L	Free-T4	1.22	ng/dL
UA	5.3	mg/dL	WT-1 mRNA	9.3×103	copies/μgRNA
CPK	65	IU/L
Na	142	mEq/L	Hematology
K	3.9	mEq/L	WBC	1,100	/μL
Cl	108	mEq/L	Neutrophil	12	%
Ca	9.6	mg/dL	Lymphocyte	86	%
P	2.6	mg/dL	Monocyte	2	%
Glucose	129	mg/dL	Myeloblast	0	%
CRP	0.09	mg/dL	RBC	275×104	/μL
PT-INR	0.97		Hemoglobin	9.6	g/dL
APTT	27.4	sec	MCV	102.9	fL
Fibrinogen	300	mg/dL	Platelet	5.3×104	/μL
FDP	<5.0	μg/mL	Reticulocyte	11	‰
D-dimer	0.9	μg/mL

CRP: C-reactive protein, UIBC: unsaturated iron-binging capacity, TSH: thyroid stimulating hormone, WT-1: Wilm’s tumor 1, WBC: white blood cell, RBC: red blood cell, MCV: mean corpuscular volume

On day 28 of UCBT, a fluorescence in situ hybridization (FISH) analysis of the sex chromosomes and a chimerism evaluation by a short tandem repeat (STR) analysis revealed the complete donor type of the bone marrow cells and the peripheral blood cells. On day 56 of UCBT, however, a FISH analysis of the sex chromosomes revealed that 2% of all nucleated cells (ANC) in the bone marrow presented with XY signals, and a STR analysis revealed that the donor chimerism of T cells in the peripheral blood had decreased to 83.5%. We made a diagnosis of molecular relapse of the original disease and reduced the dose of the immunosuppressive agent.

Thereafter, complete donor chimerism was achieved, and hematological remission was sustained for 11 months. Progressive thrombocytopenia appeared, and the amount of Wilm’s tumor 1 (WT-1) mRNA in the peripheral blood was 2.0×104 copies/μgRNA as determined by a kit using quantitative reverse transcription-polymerase chain reaction (RT-PCR) (Otsuka Pharmaceutical Company, Tokyo, Japan; normally <50 copies/μgRNA) 11 months after UCBT (Table 2). A bone marrow examination showed normocellular marrow with 7.5% abnormal blasts, suggesting a relapse of the original disease. However, an immunophenotypic analysis by flow cytometry (FCM) was unable to differentiate the leukemic blasts in the bone marrow sample.

Table 2.

The Laboratory Data at the Onset of Donor Cell Leukemia.

Blood biochemistry			Fe	140	μg/dL
Total protein	7.0	g/dL	UIBC	111	μg/dL
Albumin	4.3	g/dL	Ferritin	417	ng/mL
BUN	21.4	mg/dL	IgG	1,359	mg/dL
Creatinine	1.47	mg/dL	IgA	249	mg/dL
AST	20	IU/L	IgM	41	mg/dL
ALT	14	IU/L
LDH	233	IU/L	Hematology
ALP	266	IU/L	WBC	6,100	/μL
UA	6.8	mg/dL	Neutrophil	49	%
CPK	54	IU/L	Lymphocyte	39	%
Na	145	mEq/L	Monocyte	9	%
K	3.9	mEq/L	Eosinophil	3	%
Cl	110	mEq/L	RBC	331×104	/μL
Ca	9.2	mg/dL	Hemoglobin	10.8	g/dL
P	3.6	mg/dL	MCV	96.7	fL
Glucose	115	mg/dL	Platelet	6.9×104	/μL
CRP	0.09	mg/dL	Reticulocyte	14	‰
WT-1 mRNA	2.0×104	copies/μgRNA

CRP: C-reactive protein, WT-1: Wilm's tumor, UIBC: unsaturated iron-binging capacity, WBC: white blood cell, RBC: red blood cell, MCV: mean corpuscular volume

The subsequent clinical course is shown in Fig. 1. The patient was treated with azacitidine at a dose of 75 mg/m2 on 7 consecutive days because we initially assumed a relapse of the original disease. However, a FISH analysis before azacitidine monotherapy showed that 100% of the bone marrow cells had XX signals; therefore, we considered this to be a case of donor-derived hematologic malignancy. On day 27 of azacitidine monotherapy, a bone marrow examination revealed hematological CR. He was followed up as an outpatient without additional treatment; however, the blasts eventually reappeared and gradually increased in the peripheral blood with the elevation of lactate dehydrogenase (LDH) levels. On day 97 of azacitidine therapy, we conducted a bone marrow biopsy because of a dry tap in the bone marrow aspirate and found that myeloperoxidase-negative blasts had invaded the bone marrow (Fig. 2). An immunophenotypic analysis by FCM showed the expression of CD7, CD21, CD33, CD34, CD38, CD56 and human leukocyte antigen (HLA)-DR on the leukemic blasts. The blasts showed a normal female karyotype in a G-banding chromosomal examination. Another FISH analysis showed that 100% of the bone marrow cells had XX signals again. We therefore made a diagnosis of myeloid/natural killer cell precursor acute leukemia (M/NKPAL) derived from donor cells. Whole-body computed tomography (CT) revealed no extramedullary lesions.

Figure 1.

The clinical course of the present patient. AZA: azacitidine, BM: bone marrow, CA: cytarabine, IDR: idarubicin, L-ASP: L-asparaginase, LDH: lactate dehydrogenase, PB: peripheral blood, WT-1: Wilm’s tumor 1

Figure 2.

May-Grünwald-Giemsa-stained (Wright-Giemsa-stained) bone marrow cells (A) and peripheral blood cells (B) at the first relapse of DCL. The bone marrow was filled with myeloperoxidase-negative blast cells and a few blast cells were found in the peripheral blood. (×1,000, Wright-Giemsa staining). DCL: donor cell leukemia

Because it was not clear whether allogeneic HSCT could significantly improve the outcome of patients with DCL and because the patient refused to undergo any transplantation, we did not plan to give him a second HSCT.

He underwent chemotherapy consisting of idarubicin (10 mg/m2 on days 1-3) and cytarabine (100 mg/m2 on days 1-7). He demonstrated prolonged neutropenia, which caused a severe infection and acute kidney injury. After a single course of intensive chemotherapy, a bone marrow examination revealed CR, and the amount of WT-1 mRNA in the peripheral blood dropped below 50 copies/μgRNA.

The patient experienced a second relapse of DCL with WT-1 mRNA 8.1×104 copies/μgRNA and the appearance of blasts in the peripheral blood 2 months after achieving CR. He underwent L-asparaginase (L-ASP) monotherapy (6,000 IU/m2 on days 1-7), which caused febrile neutropenia and severe liver damage. After L-ASP monotherapy, we re-examined the surface markers of the leukemic cells by FCM, which showed positivity for CD41, as well as CD7, CD21, CD33, CD34, CD38, CD56 and HLA-DR (Fig. 3). This result led us to diagnose his DCL as acute megakaryoblastic leukemia rather than M/NKPAL.

Figure 3.

A flow cytometric analysis of DCL at the diagnosis of acute megakaryocytic leukemia. Blast cells were positive for CD7, CD21, CD33, CD34, CD38, CD41, CD56 and HLA-DR. DCL: donor cell leukemia

The L-ASP monotherapy was unfortunately ineffective; therefore, we started azacitidine monotherapy (75 mg/m2 on days 1-5, 8, 9) again. Following a single course of azacitidine, the number of abnormal blasts and amount of WT-1 mRNA in the peripheral blood were almost unchanged, but the patient's overall condition improved with a mild decrease in LDH levels; he continued to receive treatment with azacitidine alone as an outpatient. He underwent red blood cell transfusions regularly, but no severe adverse events such as febrile neutropenia were observed during azacitidine monotherapy. He demonstrated disease progression four months after azacitidine monotherapy. He underwent low-dose combination chemotherapy and gemtuzumab ozogamicin monotherapy, but they were ineffective and caused severe myelosuppression and sepsis. He ultimately died of brain hemorrhaging 28 months after UCBT.

Discussion

Acute megakaryoblastic leukemia can develop as DCL after allogeneic HSCT. Table 3 shows the main characteristics of the original leukemia and donor cell leukemia in our report. HSCT is an effective treatment for many hematologic malignancies, but disease relapse remains a major cause of post-transplant mortality (6). DCL is de novo leukemia developing in donor-derived hematopoietic precursor cells and is a rare but severe complication after allogeneic HSCT. The incidence of DCL may have been underestimated because clinicians do not perform thorough examinations to distinguish between recipient-derived cells and donor-derived cells in cases of leukemia developing after HSCT. DCL has become well known, and a number of reports of DCL have been published in recent years (7-9). It is important to distinguish DCL from relapse of original diseases when patients develop leukemia mimicking the original diseases after allogeneic HSCT, especially in cases where a decision regarding treatment for the leukemia must be made.

Table 3.

Main Characteristics of the Original Leukemia and Donor Cell Leukemia.

	Original leukemia	Donor cell leukemia
Myeloperoxidase stain	Positive	Negative
Cell surface markers	Positive: CD13, CD34, HLA-DR Negative: CD7, CD33, CD41, CD56	Positive: CD7, CD13, CD21, CD34, CD33, CD38, CD41, CD56, HLA-DR
G-banding stain	47,XY,+1,der(1;7)(q10;p10),+8[13]/46,XY[7]	46,XX[20]
WHO classification	AML with myelodysplasia-related changes	Acute megakaryoblastic leukemia

WHO: World Health Organization, AML: acute myelogenous leukemia

In the present report, we diagnosed DCL through a FISH analysis of the sex chromosomes, although we initially suspected relapse of the original disease. Compared to relapsed disease, which almost always occurs within 2 years post-transplant, DCL is commonly diagnosed later, with a median time to DCL of 31 months (range: 2-312 months) (10). Shiozaki et al. reported that the characteristics of DCL occurring after UCBT were different from those occurring after bone marrow transplantation (BMT), including the duration of time between transplantation and the occurrence of DCL, and the types of abnormal karyotypes (11). DCL occurred within a significantly shorter period after UCBT than after BMT, with a median duration of 14.5 months for UCBT and 36 months for BMT. The frequency of monosomy 7 observed in DCL after UCBT was significantly higher than that in DCL after BMT. Greaves reported that up to 5% of CB samples might harbor potentially pre-leukemic clones, raising the possibility that DCL might be disproportionately common after UCBT (12). It has been reported that risk factors for DCL include the use of cells obtained from elderly donors and cells obtained from cord blood (13). In our case, the use of cord blood for the cell source was a risk factor for developing DCL as well as for the early occurrence of DCL.

Azacitidine may be effective for treating acute megakaryoblastic leukemia. Acute megakaryoblastic leukemia is a rare form of adult AML, occurring in about 1% of all AML cases (14). Because of its low incidence, clinical data on acute megakaryoblastic leukemia are limited.

We treated the present patient with anthracycline-based chemotherapy, which led to CR. However, such an intensive chemotherapy regimen caused severe complications, including prolonged neutropenia, severe infection, acute kidney injury and severe liver damage. We next treated the patient with L-ASP alone because of our initial diagnosis of M/NKPAL, with reference to the reports of two pediatric cases of M/NKPAL successfully treated with L-ASP-based therapy (15,16). However, we reached a correct diagnosis of acute megakaryoblastic leukemia by re-assessing surface markers of his DCL including CD41, after unsuccessful treatment with L-ASP. M/NKPAL is negative for the cytochemical myeloperoxidase (MPO) reaction, suggesting that this leukemia is incorporated into the FAB classification of AML as M0. It has been shown that some cases of CD7+ CD56+ AML with MPO-negative were categorized as M7 because of CD41 positivity and infrequently showed extramedullary involvement (17). We need to discriminate between these types of leukemia by examining surface markers, including CD41, when a patient is diagnosed with CD7+ CD56+ AML. We subsequently treated the patients with azacitidine, which is effective in patients with high-risk MDS or AML who are unfit for intensive chemotherapy (4,15,18). Azacitidine actually showed some degree of efficacy in our patient.

In conclusion, DCL should be considered in all cases of acute leukemia developing after allogeneic HSCT and should be differentiated from relapse of the original disease by a thorough examination with a chimerism analysis using molecular or cytogenetic technique in cases with leukemia mimicking the original disease. Azacitidine monotherapy may be an effective alternative therapy for patients with DCL of myeloid lineage who are unsuited for intensive chemotherapy because of various complications or organ damage after HSCT. Further reports should be accumulated to determine whether or not azacitidine is as effective against de novo acute megakaryoblastic leukemia as it is for the treatment of other types of AML.

The authors state that they have no Conflict of Interest (COI).