hematology - JPLSG

9 downloads 3786 Views 494KB Size Report
(AML-TLD) [1], is defined as acute leukemia with dysplasia in 2 or more cell .... corresponding event-free survival rates at 3 years for the patients treated with ...
International Journal of

HEMATOLOGY

Acute Myeloid Leukemia with Multilineage Dysplasia in Children Souichi Adachi,a Atsushi Manabe,b Masue Imaizumi,c Takashi Taga,d Akio Tawa,e Masahito Tsurusawa,f Akira Kikuchi,g Atsuko Masunaga,h Masahiro Tsuchida,i Tatsutoshi Nakahata,a MDS Committee of the Japanese Society of Pediatric Hematology a

Department of Pediatrics, Kyoto University Graduate School of Medicine, Kyoto, Japan; bDepartment of Pediatrics, St. Luke’s International Hospital, Tokyo, Japan; cDepartment of Hematology and Oncology, Miyagi Children’s Hospital, Miyagi, Japan; dDepartment of Pediatrics, Shiga University of Medical Science, Shiga, Japan; eDepartment of Pediatrics, National Hospital Organization Osaka National Hospital, Osaka, Japan; fDepartment of Pediatrics, Aichi Medical University, Aichi, Japan; gDivision of Hematology/Oncology, Saitama Children’s Medical Center, Saitama, Japan; hDepartment of Surgical Pathology, Showa University Fujigaoka Hospital, Kanagawa, Japan; iDepartment of Pediatrics, Ibaraki Children’s Hospital, Ibaraki, Japan Received February 14, 2007; received in revised form July 20, 2007; accepted July 25, 2007

Abstract We retrospectively surveyed pediatric acute myeloid leukemia (AML) patients with multilineage dysplasia treated with the AML 99 and the Children’s Cancer and Leukemia Study Group (CCLSG) AML 9805 protocols. We found only 9 AML patients (2.6%) with multilineage dysplasia among the 341 patients with newly diagnosed de novo AML. Eight of the 9 patients obtained complete remission (CR) following the intensive AML-oriented treatments. Three of 7 patients who underwent stem cell transplantation were alive in CR for more than 4 years, and the 2 patients treated only with chemotherapy were alive in CR for more than 30 months. We did not identify any particular chromosomal abnormalities or differentiation according to the French-American-British classification in these 9 patients. No reports have described AML with multilineage dysplasia in children, and the incidence of the disease is expected to be very low. We plan to conduct a prospective pathologic review to select cases with this disease entity in the next Japanese Pediatric Leukemia/Lymphoma Study Group (JPLSG) AML-05 protocol. Int J Hematol. 2007;86:xxx-xxx. doi: 10.1532/IJH97.07025 © 2007 The Japanese Society of Hematology Key words: AML with multilineage dysplasia; Children; Retrospective analysis

improves the outcome prospects for adult AML-TLD [6]. To our knowledge, there has been no report of AML with multilineage dysplasia in pediatric patients, and the prognosis for these patients has not been determined. In fact, pediatric myelodysplastic syndromes are relatively rare, and we first reported a large-scale retrospective analysis of 189 patients in Japan in 2001 [7]. In this study, we retrospectively analyzed Japanese pediatric AML patients to define the impacts of dysplasia on the prognosis of pediatric AML.

1. Introduction Acute myeloid leukemia (AML) with multilineage dysplasia, formerly reported as AML with trilineage dysplasia (AML-TLD) [1], is defined as acute leukemia with dysplasia in 2 or more cell lines. Dysplasia must be present in more than 50% of the cells for at least 2 cell lines, including megakaryocytes [2]. The prognosis for adult AML-TLD is usually poor despite treatment with intensive chemotherapy [3-5], although allogeneic bone marrow transplantation (BMT)

2. Patients and Methods 2.1. Patients

Correspondence and reprint requests: Tatsutoshi Nakahata, Department of Pediatrics, Graduate School of Medicine, Kyoto University, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan; 81-75-751-3290; fax: 81-75-752-2361 (e-mail: [email protected]).

From January 2000 to December 2002, 240 pediatric patients with newly diagnosed de novo AML were registered for the AML 99 [8] protocol, and 101 patients were registered 1

2

Adachi et al / International Journal of Hematology 86 (2007) xx-xx

Figure 1. Schema of the AML 99 protocol. Induction C regimen: 500 mg/m2 cytosine arabinoside (Ara-C) continuously administered intravenously (div) on days 1 to 3 and 500 mg/m2 div over 20 hours on days 8 to 10; 8 mg/m2 idarubicin div over 1 hour on days 1 to 3; 200 mg/m2 etoposide (VP-16) div over 2 hours on days 8 to 10. Course 1: 3 g/m2 Ara-C div over 3 hours twice on days 1 to 3; 100 mg/m2 VP-16 div over 2 hours on days 1 to 5; 10 mg/m2 idarubicin div over 1 hour on day 1. Course 2: 200 mg/m2 Ara-C div continuously on days 4 to 8; 150 mg/m2 VP-16 div over 2 hours on days 1 to 3; 5 mg/m2 mitoxantrone div over 1 hour on days 4 to 6. Course 3: 2 g/m2 Ara-C div over 3 hours twice on days 1 to 5; 100 mg/m2 VP-16 div over 2 hours on days 1 to 5. Course 4: same as course 1. Course 5L: same as course 3. Course 5H: same as course 2. Course 6: 500 mg/m2 Ara-C div over 20 hours on days 1 to 3 and days 8 to 10; 500 mg/m2 VP-16 div over 2 hours on days 1 to 3 and days 8 to 10. FAB indicates French-American-British classification; WBC, white blood cell; CR, complete remission; Ph1, Philadelphia chromosome; allo BMT, allogeneic bone marrow transplantation; PBSC, peripheral blood stem cell; auto, autologous.

from May 1998 to November 2002 for the Children’s Cancer and Leukemia Study Group (CCLSG) AML 9805 protocol. Bone marrow (BM) morphology was examined at each institution, and BM specimens for cases with a dysplastic morphology were sent to the central review system of the MDS Committee of the Japanese Society of Pediatric Hematology. The diagnosis of AML with multilineage dysplasia was made according to the World Health Organization classification.

2.2. Chemotherapy Patients were treated with the AML 99 protocol (Figure 1), the CCLSG AML 9805 protocol (Figure 2), or the MDS 99 protocol. The AML 99 remission-induction therapy protocol (remission A) was administered to the low-risk and intermediate-risk patients and included etoposide (150 mg/m2 on days 1-5), mitoxantrone (5 mg/m2 on days 6-10), and cytosine arabinoside (Ara-C) (200 mg/m2 on days 6-12).

The high-risk group (white blood cell counts at diagnosis >100,000/μL and patients older than 2 years) received the remission-induction therapy (remission B), consisting of etoposide (100 mg/m2 on days 1-3 and days 11-13), idarubicin (8 mg/m2 on days 4-6), and Ara-C (100 mg/m2 on days 4-6 and 200 mg/m2 on days 11-13). After achieving complete remission (CR), patients received either 5 courses of consolidation therapy (the low-risk group and the intermediate-risk group without matched sibling donors) or 2 courses of consolidation therapy and stem cell transplantation (SCT) (the high-risk group and the intermediate-risk group with matched sibling donors). The CCLSG AML 9805 protocol includes 2 courses of remission-induction therapy consisting of Ara-C (160 mg/m2 on days 1-7), pirarubicin (40 mg/m2 on days 2 and 4), and vincristine (1 mg/m2 on day 7). After achieving CR, the patients received 6 courses of consolidation therapy. The MDS 99 protocol was designed for pediatric patients with refractory anemia (RA), RA with ringed sideroblasts,

AML with Multilineage Dysplasia in Children

3

Figure 2. Schema of the CCLSG AML 9805 protocol. MC: 160 mg/m2 cytosine arabinoside (Ara-C) administered intravenously (div) continuously on days 1 to 5; 5 mg/m2 mitoxantrone div over 1 hour on days 2 to 4. EC: 2 g/m2 Ara-C div over 2 hours twice on days 1 to 3; 100 mg/m2 etoposide (VP-16) div over 2 hours on days 2 to 4. AVC 2: 160 mg/m2 Ara-C div continuously on days 1 to 5; 50 mg/m2 pirarubicin (THP-ADR) div over 1 hour on day 2; 1.5 mg/m2 vincristine (VCR) div on day 5. VVEC: 200 mg/m2 VP-16 div over 2 hours on days 2 to 4; 1 g/m2 Ara-C div over 2 hours twice on days 1 to 4; 0.8 mg/m2 VCR div on day 7; 2 mg/m2 vindesine div on day 9. EMC: 10 mg/m2 mitoxantrone div over 1 hour on days 2 to 3; 100 mg/m2 VP-16 div over 2 hours on days 8 to 10; 200 mg/m2 Ara-C div continuously on days 1 to 3 and 500 mg/m2 div continuously on days 8 to 10. ci indicates continuous infusion; TIT, triple intrathecal administrations of methotrexate, Ara-C, and high-dose chemotherapy; CR, complete remission; PR, partial remission; NR, no remission; IDA, idarubicin; DEX, dexamethasone.

4

Adachi et al / International Journal of Hematology 86 (2007) xx-xx

Table 1. Characteristics of Multilineage Acute Myeloid Leukemia in Children* Case No. 1 2 3 4 5 6 7 8 9

RBC,

Age, y

Sex

2 15 7 12 4 7 15 6 10

Female Female Female Male Male Male Female Female Male

×106/μL 3.41 0.95 2.99 3.29 1.85 1.80 1.54 2.24 2.79

Plt,

×103/μL 37 18 45 115 65 12 80 41 91

WBC, /μL

PB Blasts, %

BM NCC, ×104/μL

BM Blasts, %

Karyotype

4900 1900 2750 1900 4100 3100 18,740 6100 9000

53 3 0 6 2.5 7 30 5 14

6 3.8 17.2 8.1 65 3.8 11 46.3 45

18 20 20 22.4 27 36.4 60 80 89.4

48,XX,t(7;12)(q36;p11),+8,+19 –7 Normal 47,+X add(6),+X,+19 Normal 46,XX,inv(9)(p11q13) Normal Normal

*RBC indicates red blood cells; Plt, platelets; WBC, white blood cells; PB, peripheral blood; BM, bone marrow; NCC, nucleated cell count.

RA with excess of blasts (RAEB), RAEB in transformation (RAEB-t), or juvenile myelomonocytic leukemia. According to the protocol, RAEB-t patients underwent SCT after 2 courses of chemotherapy (the same as the first 2 chemotherapy courses in the AML 99 protocol). The mean rates (±SD) of overall survival at 3 years for the 240 patients treated with the AML 99 protocol and the 45 patients treated with the CCLSG AML 9805 protocol were 79.2% ± 5.2% and 87.6% ± 6.0%, respectively. The corresponding event-free survival rates at 3 years for the patients treated with these 2 protocols were 65.4% ± 6.2% and 65.6% ± 8.6%, respectively. The CR rates for those treated with these protocols were 95% and 86%, respectively. Treatment-related mortalities were low for both protocols, and only 4 of 240 patients in the AML 99 protocol and 3 of 45 patients in the CCLSG AML 9805 protocol died in CR.

2.3. Karyotyping The chromosomes of 20 or more cells were analyzed by the standard G-banding staining methods. Patients were classified into 3 risk groups (favorable-, intermediate-, and adverse-risk groups). Karyotype classification was similar to that used in other large studies [9].

3. Results Only 9 patients (excluding Down syndrome patients) of 341 patients with de novo AML received a morphologic

diagnosis of AML with multilineage dysplasia, and these patients were treated with the AML 99 protocol or the CCLSG AML 9805 protocol. Table 1 summarizes the characteristics of these patients. The patients consisted of 4 boys and 5 girls aged 2 to 14 years (median, 7 years). The counts of white blood cells in the peripheral blood of these patients ranged from 1900/μL to 18,700/μL (median, 4100/μL), and the percentages of blasts in the peripheral blood ranged from 0% to 53% (median, 6%). Blast percentages in BM specimens ranged from 18% (case 1) to 89.4% (median, 27%), and the counts of nucleated cells in the BM ranged from 38,000/μL to 650,000/μL (median, 110,000/μL). Case 1 was diagnosed as AML with multilineage dysplasia because the percentage of blasts in the peripheral blood was 53% (Table 1). We did not find any particular patterns associated with differentiation according to the French-AmericanBritish classification (M2, 3 cases; M4, 1 case; M6, 1 case; M7, 1 case; not determined, 3 cases). Eight cases were classified cytogenetically into the intermediate-risk group (normal karyotype, 4 cases; 48,XX,t(7;12)(q36;p11),+8,+19, 1 case; 48, add(6)+X,+19, 1 case; 47,+X, 1 case; 46,XX,inv(9)(p11q13), 1 case), and 1 case was classified in the adverse-risk group (monosomy 7). All patients showed dysplasia in 3 cell lines. Representative morphologies are shown in Figure 3 (myeloid dysplasia in case 7), Figure 4 (erythroid dysplasia and myeloid dysplasia in case 5), and Figure 5 (dysplasia in megakaryocytes in case 9). The treatments and outcomes for

Figure 4. Figure 3. Myeloblast and myeloid dysplasia in case 7.

Erythroid dysplasia (arrows) and myeloid dysplasia (arrowhead) in case 5.

AML with Multilineage Dysplasia in Children

5

chemotherapy. The conditioning regimen was myeloablative in all cases: 6 patients received total body irradiation-based regimens, and 1 patient received the combination of busulfan, cyclophosphamide, and etoposide. SCT donor sources were as follows: unrelated cord blood, 3 cases; mismatched related BM, 2 cases; and unrelated BM, 2 cases. Three of the 7 transplantation patients died after SCT, and the 2 patients who did not undergo SCT survived for more than 2 years after intensive chemotherapy.

4. Discussion

Figure 5. Micromegakaryocyte (arrow) in case 9. these patients are summarized in Table 2. All of the patients were treated with intensive chemotherapeutic regimens (AML 99 protocol, 5 cases; CCLSG AML 9805 protocol, 2 cases; MDS 99 protocol, 2 cases). Eight patients achieved CR, and 7 patients underwent allogeneic SCT. The patients in case 1 and case 7 did not undergo SCT because the morphologic dysplasia disappeared after intensive remission-induction

Table 2. Characteristics of Cells of Multilineage Acute Myeloid Leukemia in Children* Case No.

FAB

Dysplasia

1

ND

2

M7

3

M2

4

ND

5

M6

6

M2

7

M4

8

M2

9

ND

My: hyper, hypo-gra, pel-Hu; E: megalo, multinuc; Me: micromega, giant plt My: hyper, hypo-gra, pel-Hu; E: megalo, multinuc; Me: micromega, giant plt My: hyper, hypo-gra, pel-Hu; E: megalo; Me: micromega, multinuc mega My: hypo-gra, pel-Hu; E: megalo, multinuc; Me: micromega, multinuc mega My: hyper, pel-Hu; E: megalo, multinuc; Me: micromega My: hyper, hypo-gra; E: megalo, multinuc; Me: micromega My: hyper, pel-Hu; E: megalo, multinuc; Me: micromega, giant plt My; hypo-gra, pel-Hu; E: megalo; Me: micromega My: hyper, hypo-gra, pel-Hu; E: megalo; Me: micromega

*FAB indicates French-American-British classification; ND, not determined; My, myeloid; hyper, neutrophils with hypersegmented nuclei; hypo-gra, hypogranulated neutrophils; pel-Hu, pseudo-Pelger-Huet anomaly; E, erythroid; megalo, megaloblastoid change; multinuc, multinucleated erythroid cell; Me, megakaryocytic; micromega, micromegakaryocyte; multinuc mega, multinucleated megakaryocyte; giant plt, giant platelets.

This report is the first of surveillance of pediatric AML patients with multilineage dysplasia according to the World Health Organization classification [9]. We found only 9 AML patients with multilineage dysplasia (2.6%) among 341 patients with newly diagnosed de novo AML. Because the assessment of dysplastic features in children was difficult, we also conducted a central review of pathologic specimens. In fact, the percentages of blasts in the peripheral blood and BM determined by the original institution and by the central review were quite different in some cases (5% versus 53% peripheral blood blasts in case 1 and 7.6% versus 27% BM blasts in case 5). This finding underscores the importance of central review in cases of pediatric AML-TLD. We may have overlooked some AML-TLD cases because the system of central review was not part of either the AML 99 study or the CCLSG AML 9805 study. We suggest, however, that pediatric AML-TLD is very rare, as we expected. TLD frequencies between 7.8% and 27.7% have been reported in adult cases [1-4, 10-16]. None of the patients had favorablerisk cytogenetic patterns, and most were classified in the intermediate-risk group. We were unable to conclude from our small number of patients that dysplasia had any adverse effects on the prognosis of AML. Intensive chemotherapy may improve the prognosis for pediatric AML patients, even for patients with a dysplastic morphology, because 8 of the 9 patients achieved CR. According to the MDS 99 protocol, patients with fewer than 30% BM blasts at diagnosis should receive an SCT after 2 courses of chemotherapy, but 1 patient did not undergo SCT because the morphologic dysplasia disappeared after the remissioninduction chemotherapy. Another patient with more than 30% BM blasts also did not receive an SCT because the dysplastic features resolved after intensive chemotherapy. These 2 patients did not receive an SCT because the indication for SCT was not determined in the treatment protocol and the responses to chemotherapy were good in these patients. Still, 7 of 9 patients received an SCT after 2 or 3 courses of intensive chemotherapy, because allogeneic BMT improves the outcomes of adult AML-TLD patients [6]. Four of the 7 patients who received an SCT died, whereas both patients who did not receive an SCT survived for more than 2 years (Table 3). Their BM smears showed no dysplastic features after remission-induction chemotherapy. Most Japanese pediatric AML patients will be registered in the next Japanese Pediatric Leukemia/Lymphoma Study Group (JPLSG) AML 05 protocol, and a system of central morphologic review will be introduced into the study. Thus, we will know in the near future the exact incidence of pediatric AML with multilineage dysplasia. In the JPLSG

6

Adachi et al / International Journal of Hematology 86 (2007) xx-xx

Table 3. Treatments and Outcomes for Acute Myeloid Leukemia Patients with Multilineage Dysplasia* Patient No.

Dysplasia after Induction Tx

CR after Induction Tx

SCT

1 2 3 4 5 6

No No No No Yes No

Yes Yes Yes Yes No Yes

No Yes Yes Yes Yes Yes

7 8 9

No Yes Yes

Yes Yes Yes

No Yes Yes

SCT Donor Source

Conditioning Regimen

GVHD Prophylaxis

EFS, mo

OS, mo

Outcome

31+ 22 69+ 49+ 15 71+

Alive Dead (DOD) Alive Alive Dead (DOD) Alive

31+ 18 21

Alive Dead (GVHD) Dead (DOD)

UBMT UBMT UCBT UCBT RBMT (mismatched)

TBI + CY TBI + L-PAM TBI + CY BU + VP + CY TBI + Ttepa + L-PAM

MTX + FK MTX + FK FK + mPSL CyA + MTX CyA + MTX

31+ 10 69+ 49+ 0 71+

UCBT RBMT (mismatched)

TBI + Ara-C + CY TBI + VP + L-PAM

CyA + MTX FK + MTX

31+ 18 14

*Tx indicates therapy; CR, complete remission; SCT, stem cell transplantation; GVHD, graft-versus-host disease; EFS, event-free survival; OS, overall survival; UBMT, unrelated bone marrow transplantation; TBI, total body irradiation; CY, cyclophosphamide; MTX, methotrexate; FK, tacrolimus; DOD, death of disease; L-PAM, melphalan; UCBT, unrelated cord blood transplantation; mPSL, methylprednisolone; Bu, busulfan; VP, etoposide; CyA, cyclosporin A; RBMT, related bone marrow transplantation; Ttepa, thiotepa; Ara-C, cytosine arabinoside.

AML 05 protocol, only high-risk AML patients (ie, with monosomy 7, Philadelphia chromosome positivity, or non-CR after induction therapy) will receive an SCT during the first CR, because intensive chemotherapy was shown in the previous AML 99 study to improve the prognosis of pediatric AML patients [17]. At present, this protocol should be employed even in patients with dysplastic morphologies, because morphologic dysplasia has no independent prognostic relevance in the context of intensive chemotherapy. In conclusion, this report is the first of pediatric AML with multilineage dysplasia. We found that the clinical features of the disease were heterogeneous and that the disease in children may be different from that in adults with respect to the prognosis. We are currently conducting a clinical trial to further characterize the disease in a prospective manner.

Acknowledgments The authors express sincere gratitude to the members of the Japan Society of Pediatric Hematology for their cooperation in this study. Drs. Arata Watanabe, Kazuko Hamamoto, Rie Kanai, Yasushi Noguchi, Ken Tabuchi, Yuko Cho, and Yasuzo Hirota provided the patient data.

References 1. Brito-Babapulle F, Catovsky D, Galton DA. Clinical and laboratory features of de novo acute myeloid leukemia with trilineage myelodysplasia. Br J Haematol. 1987;66:445-450. 2. Gahn B, Haase D, Unterhalt M, et al. De novo AML with dysplastic hematopoiesis: cytogenetic and prognostic significance. Leukemia. 1996;10:946-951. 3. Goasguen JE, Matsuo T, Cox C, Bennett JM. Evaluation of the dysmyelopoiesis in 336 patients with de novo acute myeloid leukemia: major importance of dysgranulopoiesis for remission and survival. Leukemia. 1992;6:520-525. 4. Kuriyama K, Tomonaga M, Matsuo T, et al. Poor response to intensive chemotherapy in de novo acute myeloid leukemia with trilineage myelodysplasia. Japan Adult Leukaemia Study Group (JALSG). Br J Haematol. 1994;86:767-773. 5. Miyazaki Y, Kuriyama K, Miyawaki S, et al, Japan Adult Leukaemia Study Group. Cytogenetic heterogeneity of acute myeloid leukemia (AML) with trilineage dysplasia: Japan Adult Leukaemia Study Group-AML 92 study. Br J Haematol. 2003;120:56-62.

6. Taguchi J, Miyazaki Y, Yoshida S, et al. Allogeneic bone marrow transplantation improves the outcome of de novo AML with trilineage dysplasia (AML-TLD). Leukemia. 2000;14:1861-1866. 7. Sasaki H, Manabe A, Kojima S, et al. Myelodysplastic syndrome in childhood: a retrospective study of 189 patients in Japan. Leukemia. 2001;15:1713-1720. 8. Tsukimoto I, Tawa A, Hanada R, et al. Excellent outcome of risk stratified treatment for childhood acute myeloid leukemia-AML 99 trial: for the Japanese Childhood AML Cooperative Study Group. Blood. 2005;106:261a. Abstract 889. 9. Grimwade D, Walker H, Oliver F, et al. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial.The Medical Research Council Adult and Children’s Leukaemia Working Parties. Blood. 1998;92:2322-2333. 10. Brunning RD, Matutes E, Harris NL, et al. Acute myeloid leukemia with multilineage dysplasia. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. World Health Organization Classification of Tumours. Pathology & Genetics: Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001. 11. Estienne MH, Fenaux P, Preudhomme C, et al. Prognostic value of dysmyelopoietic features in de novo acute myeloid leukemia: a report on 132 patients. Clin Lab Haematol. 1990;12:57-65. 12. Ballen KK, Gilliland DG, Kalish LA, Shulman LN. Bone marrow dysplasia in patients with newly diagnosed acute myelogenous leukemia does not correlate with history of myelodysplasia or with remission rate and survival. Cancer. 1994;73:314-321. 13. Meckenstock G, Aul C, Hildebrandt B, et al. Dyshematopoiesis in de novo acute myeloid leukemia: cell biological features and prognostic significance. Leuk Lymphoma. 1998;29:523-531. 14. Kahl C, Florschütz A, Müller G, et al. Prognostic significance of dysplastic features of hematopoiesis in patients with de novo acute myelogenous leukemia. Ann Hematol. 1997;75:91-94. 15. Kuriyama K, Tomonaga M, Kobayashi T, et al. Morphological diagnosis of the Japan Adult Leukemia Study Group acute myeloid leukemia protocols: central review. Int J Hematol. 2001;73:93-99. 16. Haferlach T, Schoch C, Löffer H, et al. Morphologic dysplasia in de novo acute myeloid leukemia (AML) is related to unfavorable cytogenetics but has no independent prognostic relevance under the conditions of intensive induction therapy: results of a multiparameter analysis from the German AML Cooperative Group studies. J Clin Oncol. 2003;21:256-265. 17. Tomizawa D, Tabuchi K, Kinoshita A, et al. Repetitive cycles of high-dose cytarabine are effective for childhood acute myeloid leukemia: long-term outcome of the children with AML treated on two consecutive trials of Tokyo Children’s Cancer Study Group. Pediatr Blood Cancer. 2007;49:127-132.