VOLUME9MARCH2018

57

B. De Moerloose, MD, PhD1,2, E. Nauwynck, MD3, K. Arts, MD4, L. Willems, MD, PhD1, V. Labarque, MD, PhD4, T. Lammens, PhD1,2, A. Uyttebroeck, MD, PhD4

1Department of Paediatric Haematology-Oncology and Stem Cell Transplantation, Ghent University Hospital, Ghent, Belgium, 2Cancer Research

Institute Ghent (CRIG), Ghent, Belgium, 3Department of Pediatrics, Clinique Sainte-Anne Saint-Rémi, Anderlecht, Belgium, 4Department of Pae-

diatric Haematology-Oncology, University Hospital Leuven, Leuven, Belgium.

Please send all correspondence to: B. De Moerloose, MD, PhD, Afdeling pediatrische hemato-oncologie, 3K12D, Universitair Ziekenhuis Gent,

Corneel Heymanslaan 10, B-9000 Gent, tel: +32 9 332 24 16, email: barbara.demoerloose@uzgent.be.

Conflict of interest: The authors have nothing to disclose and indicate no potential conflict of interest.

Keywords: acute leukaemia, ALL, AML, infant, outcome, treatment.

INTRODUCTIONLeukaemia occurring in the first year of life is a very rare disease, with only a few infant cases diagnosed each year in specialised paediatric oncological centres. According to the Belgian Cancer Registry, the age specific incidence for infant leukaemia was 34 per million per year in the period 2004-2009, and the overall survival (OS) rate for these children was 44%.1 It is, however, the third most common cancer in infants in Belgium, after neuroblastoma and Central Nervous System (CNS) tumours. Half of Belgian infants with leukae-mia are diagnosed with acute lymphoblastic leukaemia (ALL),

42% suffer from acute myeloid leukaemia (AML) and the remaining cases belong to other rare subtypes of leukaemia, such as juvenile myelomonocytic leukaemia.1 There is a slight female predominance in this infant disease.2,3 In ALL, the lymphoblasts mostly originate from the B-cell line.4 The most common French-American-British (FAB) subtypes in infants with AML are M4 (acute myelomonocytic leukaemia), M5 (acute monoblastic leukaemia) and M7 (acute megakaryo-blastic leukaemia).5,6 The exact cause of infant leukaemia remains unknown, but there are some studies which demonstrate that probably all

SUMMARYInfant leukaemia is a rare disease but the 3rd most frequent malignancy in this age group. Both acute lympho-blastic leukaemia and acute myeloid leukaemia in the first year of life have particular clinical and biological characteristics such as B-cell phenotype with co-expression of myeloid markers in acute lymphoblastic leukaemia, FAB M5 or M7 in acute myeloid leukaemia, the presence of extramedullary symptoms and a high frequency of KMT2A rearrangements. Survival rates for infant acute leukaemia are worse than for older children. In this study, the characteristics and outcome of 50 infants with acute lymphoblastic leukaemia and acute myeloid leukaemia treated at the University Hospitals of Ghent and Leuven between 1989 and 2015 were studied and correlated with literature data. With event-free survival and overall survival rates of 44% and 52% for the entire cohort, the outcome of these patients was comparable to those in published clinical trials. In general, the event-free survival and overall survival was superior in acute myeloid leukaemia compared to acute lymphoblastic leukaemia infants and not influenced by age (< or ≥6 months), white blood cell count at diagnosis or presence of a KMT2A rearrangement. For future trials in infant leukaemia, the high number of early deaths, toxic deaths and relapses remain the most challenging problems.(BELG J HEMATOL 2018;9(2):57-63)

Retrospective study of infant leukaemia in the University Hospitals of Leuven and Ghent

VOLUME9MARCH2018

REVIEW HEMATOLOGY

2

58

infant leukaemia’s originate in utero due to prenatal leukae-mogenic events such as genetic rearrangements.2,7-9 In infant leukaemia, the most common genetic rearrangement is the

mixed lineage leukaemia gene (MLL, also called KMT2A) rearrangement which can be found in 75-80% and 50-60% of the ALL and AML infants, respectively.3,6,8,10 The most

TABLE 1. Patient and leukaemia characteristics at diagnosis.

ALL AML P-Value

Total number of patients 25 25

Age (days)

Median (range) 144 (1-366) 193 (1-342) P = 0.530

<6m 16 (64%) 12 (48%) P = 0.393

6-12m 9 (36%) 13 (52%)

Gender P = 0.769

M 8 (32%) 10 (40%)

F 17 (68%) 15 (60%)

Leukocytes P = 0.004

Median ( x109/L) 178 18.8

Range ( x109/L) 2.41-1 498 2.09-687.08

CNS leukaemia P = 0.156

CNS-1 11 (44%) 17 (68%)

CNS-2 2 (8%) 2 (8%)

CNS-3 9 (36%) 3 (12%)

Unknown 3 (12%) 3 (12%)

ALL type

B-ALL 23 n.a.

T-ALL 1 n.a.

AML FAB class

M5 n.a 13

M7 n.a. 5

MPAL 1 1

M: male; F: female; m: months; CNS-1: no blasts in the CSF and WBC less than 5 WBC/μL; CNS-2: presence of blasts but WBC less than 5 WBC/μL; CNS-3: presence of blasts and more than 5 WBC/μL; MPAL: mixed phenotype leukaemia; n.a.: not applicable.

VOLUME9MARCH2018

common partner chromosomes for this reciprocal transloca-tion are chromosome 4, 9 and 19 in ALL, and chromosome 9, 10 and 19 in AML.2,3,10,11

Compared to older children with acute leukaemia, infants suffer from a more severe clinical disease course. Hepato-splenomegaly, hyperleukocytosis and skin infiltrates, also known as leukaemia cutis, are commonly found.3,12 Whereas infants with AML are treated with the same treat-ment protocols as older children, there are specific infant ALL protocols since + 15 years.13-15 These studies, i.e. Interfant 99 and Interfant 06, are hybrid treatment protocols with elements designed to treat both ALL and AML and to iden-tify prognostic factors for outcome. Despite the development of these improved treatment protocols, infants with ALL only have an average survival of 40-50%.10-12 An important problem is the fact that despite achieving complete remission in the majority of ALL infants, there is a high relapse rate.13,14 Infants with AML have a similar survival as the older AML children, which is between 44-72%. As such, there is a diffe-rence in prognosis for AML or ALL infants, with AML infants having a better prognosis than ALL infants.12 The most important prognostic factors in infant ALL are younger age at diagnosis, high white blood cell (WBC) count, invasion of the CNS, presence of KMT2A rearrangements and a poor early response to therapy.3,10,13-15 On the contrary, prognostic factors for AML infants are not yet defined. The aim of this study was to evaluate all infant cases, treated for acute leukaemia in the University Hospitals of Ghent and Leuven from 1989 to 2015, to get a detailed description of this patient population, treatment results and a comparison with literature data.

PATIENTS AND METHODSAfter approval by the local ethical boards and obtaining informed consent, the files from all infant leukaemia patients treated either in the University Hospital Gasthuisberg Leuven or in the Ghent University Hospital were checked for patient and leukaemia characteristics (age, gender, leu-kaemia type, symptoms, blood and bone marrow results, cytogenetic characteristics), response to treatment and out-come (event-free survival (EFS) and OS). Events were defined as early death, no remission (after the induction phase for ALL patients, and after two induction courses for AML patients), relapse or death in complete remission. Over the years, patients were treated according to different treatment protocols with EORTC-58881, Interfant 99 and Interfant 06 for ALL cases, and EORTC-58921, ELAM02, DB AML-01 and NOPHO-DBH AML 2012 for infants with AML.14-19

Characteristics in the ALL and in the AML group were com-pared using a student’s T-test or Fisher’s Exact test with P-value=0,05 as level of significance (two-sided). Predictors of response at day 8 for ALL infants were evaluated using logistic regression analysis. To compare EFS and OS between the two leukaemia groups Kaplan-Meier graphs were genera-ted and a Logrank test was applied to demonstrate possible differences in survival between AML and ALL. For estimates of EFS, an event was defined as failure to achieve complete remission, induction death, relapse, development of a second malignancy, or death due to any cause, whichever occurred first. EFS was calculated from date of diagnosis to the date of first event. OS was calculated from date of diagnosis to the date of last follow-up or time of death due to any cause.

FIGURE 1. Distribution of leukocyte count at diagnosis for infant ALL and AML.

59

REVIEW HEMATOLOGY

VOLUME9MARCH20182

60

Predictors of EFS and OS were evaluated by Cox propor- tional hazard survival analysis. The statistical analysis was done using the program Statistical Package for Social Sciences 24 (SPSS 24).

RESULTSIn the previous 26 years (1989-2015), 50 infants were treated for acute leukaemia in Ghent and Leuven: 25 infants with ALL and 25 with AML. All but one ALL patient were of precursor-B cell origin and, as expected, 13/25 AML cases were FAB M5 and 5/25 FAB M7. One child with mixed phenotype acute leukaemia (MPAL) was treated as ALL, and one as AML. Two neonates with AML had Down syndrome. Other patient and disease characteristics are listed in Table 1. The median age at presentation was somewhat younger for the ALL group (four months) compared to the AML group (six months) and there was a female predominance in both groups. The median WBC count at diagnosis was signifi-cantly higher in ALL than in AML with WBC ranging from 2.41 to 1498 x109/L in ALL and 2.09 to 687 x109/L in AML (p=0.004). Significant differences (p<0.001) in the leuko-cyte count distribution were observed between ALL and AML infants with 19/25 (76%) ALL patients harbouring a leukocyte count of more than 50 x109/L. Most of the AML infants (21/25, 84%) had a leukocyte count ≤ 50 x109/L (Figure 1). Other blood values including thrombocytes, haemoglobin and LDH levels, and bone marrow results were not significantly different between the two patient groups. CNS involvement, i.e. CNS-3 status, was found in one third of all ALL patients (36%) whereas this was only the case for three AML cases. Hepatomegaly and splenomegaly were present in 96% and 88% of infants with ALL, respectively, and this was significantly more frequent than in the AML group. On the other hand, almost half of AML infants (12/25) presented with leukaemia cutis. Cytogenetic and molecular characteristics were studied by karyotyping, Fluorescence In Situ Hybridisation (FISH) and

Polymerase Chain Reaction (PCR), but FISH and PCR results were frequently lacking in the first decade of the study period. The PCR technique to detect a KMT2A rearrangement is used since the second half of the nineties. Nowadays a multiplex PCR is performed, where various targets can be examined simultaneously.Of the 50 infants with acute leukaemia, 28 had a KMT2A rearrangement: 16/25 (64%) ALL patients and 12/25 (48%) AML patients (Table 2). There was no statistical significant difference (p=0.254) in the presence of a KMT2A rearran- gement between the two patient groups, although we did observe this genetic abnormality more frequently in ALL infants. The respective partner chromosomes are shown in Table 3: chromosome 4 and chromosome 9 were the most frequent partner chromosome in ALL and AML cases, res-pectively. Of the thirteen AML patients without a KMT2A rearrangement, five had a normal karyotype, four had a translocation t(1;22), two had a complex karyotype, one had a deletion of chromosome 11 and one patient had a translo-cation t(8;16). Of the nine ALL patients without a KMT2A rearrangement, four had a normal karyotype and four had a complex karyotype. One infant with ALL had a normal karyotype at diagnosis, but presented with a KMT2A rear-rangement with chromosome 9 at relapse. It is possible that the KMT2A rearrangement was missed at initial presentation, because there was no PCR performed to confirm the results. We also investigated the possible relation between early res-ponse on day 8 in ALL patients, the leukocyte count at diag-nosis, age at presentation, the presence of a KMT2A gene rearrangement and the therapy protocol used. Each of these comparisons showed a non-significant result, which could be due to the low number of patients.The 5-year EFS and OS for the total cohort was 44% and 52%, respectively. As shown in Figure 2, the 5-year EFS was higher in infants with AML (50%) than ALL (35%), but not significantly different. Remarkably, all events occurred in the first three years after diagnosis: twelve events in AML

TABLE 2. Number of ALL and AML patients harbouring a KMT2A rearrangement.

Leukemia Type Total

ALL n (%) AML n (%)

KMT2A negative 9 (36%) 13 (52%) 22 (44%)

KMT2A positive 16 (64%) 12 (48%) 28 (66%)

Total 25 (100%) 25 (100%) 50 (100%)

VOLUME9MARCH2018

and sixteen events in ALL. There were four very early deaths in the AML patients (mostly neonates) and three in the ALL group. Five children died of toxicity while being in complete remission: a Down syndrome-AML patients had an un- expected cardiac arrest and another AML patient died of aspergillosis and multiorgan failure. In the ALL group, one child died from a pseudomonas sepsis and two from trans-plant-related toxicity (veno-occlusive disease and adeno- virus infection/graft versus host disease, respectively). Three of the six relapsed AML patients could be salvaged with second line therapy, including hematopoietic stem cell trans-plantation. From the ten ALL relapses, only one child sur- vived. This resulted in a non-significant 5-year OS difference (Figure 3, Logrank p=0.179) between the AML patients (64%) compared to infants with ALL (40%) in the study cohort. Age at diagnosis (0-6 months or 6-12 months), WBC count (<50000/µL or ≥ 50000/µL) and presence of a KMT2A rear-rangement did not influence EFS and OS (data not shown).

DISCUSSIONLeukaemia in infants is an infrequent but very challenging disease because of the high rate of toxic complications and relapses. With 50 infants diagnosed in 26 years (1989-2015) in two larger Belgian paediatric haemato-oncology centres in Belgium, this study confirms the rareness of this disorder. Infant ALL with rearranged KMT2A is recognised as a highly aggressive leukaemia with high leukocyte count and frequent involvement of extramedullary sites at diagnosis.20 As expec-ted, the median leukocyte count in our population was higher in ALL than in AML cases and organomegaly and CNS involvement were more frequently found, whereas

leukaemia cutis was more frequent in the AML cases. Now-adays, KMT2A rearrangements are found in approximately 80% of infant ALL but in our series, this was only the case for 16/25 (64%) patients, probably because some were mis-sed in the era when molecular techniques were not yet per-formed routinely.13

Only one child with infant ALL had a T-cell phenotype and this child was almost one year of age at diagnosis. All others had B-lineage ALL and frequently expression of myeloid markers (data not shown), reflecting the characteristic infant ALL phenotype being a very immature precursor-B stage with expression of CD34 and CD19 and absence of CD10 expression. These leukaemia’s seem to arise in an early progenitor cell with both B-lymphoid and myeloid characte-ristics and not fully committed to the lymphoid lineage.20 Hence, several cases of lineage switch from B-cell lineage to monocytic lineage leukaemia have been reported.21 The MPAL diagnosis in two infants in our study is probably a reflection of this leukaemia originating in a very immature progenitor cell. Current treatment protocols for infant ALL have incorporated chemotherapy usually given to AML patients, but despite these modifications, outcome for infant ALL remains poor.13,20 In comparison to older children with ALL, where EFS rates of 80.4-87.2% and OS rates of 91.8-93.5% have been reached in recently completed trials, the outcome of ALL infants is still lagging behind with EFS and OS remaining at 41.7-50.9% and 44.8-60.5%, respectively.14,20 The low survival rates of the ALL patients in our study is in line with these findings. Further intensification of the conventional chemo-therapeutic drugs with or without hematopoietic stem cell

FIGURE 3. Overall survival in infant leukaemia.FIGURE 2. Event-free survival in infant leukaemia.

Time (days)0 2000 4000 6000 8000 10000

Surv

ival

Pro

babi

lity

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

ALLAML

Surv

ival

Pro

babi

lity

Time (days)

0 2000 4000 6000 8000 10000

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

61

REVIEW HEMATOLOGY

VOLUME9MARCH20182

62

transplantation is not possible in these already heavily treated patients with high incidences of toxic complications. Other compounds such as clofarabine, Flt3 inhibitors and epigenetic modifiers are candidates for future treatment modifications since infant ALL cells are sensitive to purine nucleoside analogues, have high expression of wild-type Flt3 (especially in KMT2A-rearranged cases) and are characterised by an aberrant methylated genomic state in genome-wide methylation studies.3,20,22,23 Other treatment modalities such as immunotherapy with CD19-targeting BiTE antibodies and CAR-T therapy are currently in evaluation.20 On the contrary, the outcome of AML infants is very similar to the prognosis of older children with AML.6 Masetti et al. published an overview of the outcome of AML patients in recent paediatric AML trials and found EFS and OS rates of 44-72% and 61-76%, respectively, for children up to two years of age.24 In our study, an EFS of 51% and an OS of 64% was found for infants with AML. This correlates well with the outcome of infants in the most recent completed paediatric AML trial in Belgium, DB AML-01, in which 16/112 patients were diagnosed in the 1st year of life and a 3-year EFS of 48% (95%CI 22-70), 3-year cumulative incidence of relapse of 45% (95%CI 23-75%) and a 3-year OS of 75% (95%CI 46-90) was found for infants (personal communication).19 Four children with AML in our retrospective study died very soon after diagnosis, sometimes even before start of chemo-therapy, due to multiorgan failure, capillary leak or severe haemorrhages. Three of these four patients had congenital

leukaemia and were in poor general condition at birth. The optimal choice and dosing of chemotherapy in infants, and particularly in neonates, is a recurrent problem.25 Whereas it is common practice to reduce doses in infants, there are inconsistencies between clinical protocols and also in the magnitude of these reductions. The scientific rationale for these dose reductions is limited and pharmacokinetic studies are scarce. The physiological development of a child, showing developmental changes in distribution sites of total body water, extracellular water and body fat, the differences in avidity for binding to serum proteins and the changes in metabolic capacity and renal function during maturation, should be taken into account in future protocols designed to treat infants.26

CONCLUSIONALL and AML is a rare disease in infants with typical characteristics and therapeutic challenges. The outcome of these children, diagnosed and treated in two larger Belgian paediatric haematology-oncology sites, correlates well with overall Belgian and literature data. Because of the rarity of the disease, intensive treatment schedules and high risk of toxic complications, treatment of these cases in experienced sites is necessary. Further research for optimal dosing is absolutely warranted for future treatment protocols, as well as implementation of other modalities such as epigenetic treatment and immunotherapy in treatment protocols for infant ALL which has a worrisome outcome.

TABLE 3. KMT2A translocation partners in ALL and AML infants.

Leukemia Type Total

ALL n (%) AML* n (%)

Chromosome 1 1 (6.25%) 1 (9.09) 2 (7.41%)

Chromosome 2 1 (6.25%) 0 (0%) 1 (3.70%)

Chromosome 4 11 (68.75%) 1 (9.09) 12 (44.44%)

Chromosome 9 1 (6.25%) 5 (45.45%) 6 (22.22%)

Chromosome 10 0 (0%) 3 (27.27%) 3 (11.11%)

Chromosome 19 2 (12.5%) 1 (9.09) 3 (11.11%)

Total 16 (100%) 11 (100%) 27 (100%)

*For one patient, the translocation partner was not specified.

VOLUME9MARCH2018

REFERENCES1. Henau K, Van Damme N, Calay F, et al. Cancer incidence in Belgium (2010).

Special issue: Cancer in children and adolescents. Available from: http://www.

kankerregister.org/Cijfers_over_kanker.

2. Biondi A, Cimino G, Pieters R, et al. Biological and therapeutic aspects of

infant leukemia. Blood. 2000;96(1):24-33.

3. Brown P. Treatment of infant leukemias: challenge and promise. Hematology

Am Soc Hematol Educ Program. 2013;2013:596-600.

4. Campos-Sanchez E, Toboso-Navasa A, Romero-Camarero I, et al. Acute

lymphoblastic leukemia and developmental biology: a crucial interrelationship.

Cell Cycle. 2011;10(20):3473-86.

5. Cheng J, Sakamoto KM. Topics in paediatric leukemia – Acute myeloid

leukemia. MedGenMed. 2005;7(1):20.

6. Creutzig U, Zimmermann M, Bourquin JP, et al. Favorable outcome in infants

with AML after intensive first and second-line treatment: an AML-BFM study

group report. Leukemia. 2012;26(4):654-661.

7. Wasserman R, Galili N, Ito Y, et al. Predominance of fetal type DJH joining in

young children with B precursor lymphoblastic leukemia as evidence for an in

utero transforming event. J Exp Med. 1992;176(6):1577-81.

8. Inaba H, Greaves M, Mullighan CG. Acute lymphoblastic leukaemia. Lancet.

2013;381(9881):1943-55.

9. Gill Super HJ, Rothberg PG, Kobayashi H, et al. Clonal, nonconstitutional

rearrangements of the MLL gene in infant twins with acute lymphoblastic leukemia:

in utero chromosome rearrangement of 11q23. Blood. 1994;83(3):641-4.

10. Stam RW, den Boer ML, Pieters R. Towards targeted therapy for infant acute

lymphoblastic leukaemia. Br J Haematol. 2006;132(5):539-51.

11. Daser A, Rabbitts TH. Extending the repertoire of the mixed-lineage leukemia

gene MLL in leukemogenesis. Genes Dev. 2004;18(9):965-74.

12. van der Linden MH, Creemers S, Pieters R. Diagnosis and management of

neonatal leukaemia. Semin Fetal Neonatal Med. 2012;17(4):192-5.

13. Hilden JM, Dinndorf PA, Meerbaum SO, et al. Analysis of prognostic factors

of acute lymphoblastic leukemia in infants: report on CCG 1953 from the

Children's Oncology Group. Blood. 2006;108(2):441-51.

14. Pieters R, Schrappe M, De Lorenzo P, et al. A treatment protocol for infants

younger than 1 year with acute lymphoblastic leukaemia (Interfant-99): an obser-

vational study and a multicentre randomised trial. Lancet. 2007;370(9583):240-50.

15. Interfant06 trial. Available from: http://www.cancer.gov/clinicaltrials/search/

view?cdrid=570260&version=HealthProfessional.

16. Vilmer E, Suciu S, Ferster A, et al. Long-term results of three randomized trials

(58831, 58832, 58881) in childhood acute lymphoblastic leukemia: a CLCG-EORTC

report. Children Leukemia Cooperative Group. Leukemia. 2000;14(12):2257-66.

17. Entz-Werle N, Suciu S, van der Werff ten Bosch J, et al. Results of 58872

and 58921 trials in acute myeloblastic leukemia and relative value of chemo-

therapy vs allogeneic bone marrow transplantation in first complete remission:

the EORTC Children Leukemia Group report. Leukemia. 2005;19(12):2072-81.

18. Alloin AL, Leverger G, Dalle JH, et al. Cytogenetics and outcome of allogeneic

transplantation in first remission of acute myeloid leukemia: the French paediatric

experience. Bone Marrow Transplant. 2017;52(4):516-521.

19. de Bont E, Reedijk A, Lammens T, et al. Excellent outcome in paediatric AML

with response guided chemotherapy without allogeneic HSCT in first complete

remission: Results from protocol DB-AML01. Blood. 2015;126(23):2506.

20. Tomizawa D. Recent progress in the treatment of infant acute lymphoblastic

leukemia. Pediatr Int. 2015;57:811-819.

21.Sakaki H, Kanegane H, Nomura K, et al. Early lineage switch in an infant

acute lymphoblastic leukemia. Int J Hematol. 2009;90:653-5.

22. Garrido Castro P, van Roon EHJ, Pinhanços SS, et al. The HDAC inhibitor

panobinostat (LBH589) exerts in vivo anti-leukaemic activity against MLL-rear-

ranged acute lymphoblastic leukaemia and involves the RNF20/RNF40/WAC-

H2B ubiquitination axis. Leukemia. 2017. [Epub ahead of print].

23. Chijimatsu I, Imanaka Y, Tomizawa D, et al. Azacitidine successfully main-

tained the second remission in an infant with KMT2A-rearranged acute lympho-

blastic leukemia who relapsed after unrelated cord blood transplantation. Pediatr

Blood Cancer. 2017;64(12).

24. Masetti R, Vendemini F, Zama D, et al. Acute myeloid leukemia in infants:

biology and treatment. Front Pediatr. 2015;3(37):1-7.

25. Pieters R. Infant acute lymphoblastic leukemia: lessons learned and future

directions. Curr Hematol Malig Rep. 2009;4:167-74.

26. Kearns GL, Abdel-Rahman SM, Alander SW, et al. Developmental pharma-

cology – drug disposition, action, and therapy in infants and children. N Engl J

Med. 2003;349(12):1157-67.

KEY MESSAGES FOR CLINICAL PRACTICE

1 Infant ALL and AML are the 3rd most frequent malignancy in this age category.

2 The outcome of infant ALL is inferior to the outcome of ALL in older age categories, whereas infant AML cure rates are comparable to these of older children.

3 Because of the rarity of the disease, the complexity of treatment protocols and clinical management, and the high rate of toxic complications, treatment of these patients should be reserved to experienced sites.

4 For further outcome improvement, particularly in infant ALL, new treatment modalities should be incorporated into international treatment protocols.

63 REVIEW HEMATOLOGY