|
Introduction
Acute lymphoblastic leukemia (ALL),
the most common childhood cancer, has been an emblem of medical
progress, with steady improvement in cure rate over the past 50
years and a current 5 year event-free survival rate of
approximately 80 percent. An understanding of molecular genetics is
playing an increasingly important role in optimizing therapy in
pediatric ALL, defining distinct prognostic subgroups for which
therapy can be tailored so that low-risk patients are spared
unnecessary toxicity, while high-risk patients receive the
intensive therapy most likely to effect a cure.
Prognostic factors
Age and initial white blood count (WBC) are two powerful and
independent parameters used to guide initial therapy. The
Rome-NCI consensus criteria define high-risk ALL as older than 1 year
or younger than 10 years and/or initial WBC greater than 50,000/ml). Immunophenotype
also plays an important role in determining choice of
therapeutic regimen. A wealth of additional information is
available at the molecular level regarding disease etiology,
natural history and prognosis. The major prognostically
important recurrent abnormalities in pediatric ALL are reviewed
in the following sections.
Overview of ALL molecular genetics
Cytogenetics
Cytogenetic analysis of karyotype, more recently
supplemented by a panel of specific fluorescent in-situ
hybridization (FISH) probes, is a crucial element in ALL
diagnostic evaluation. Current cytogenetics laboratories
detect karyotypic abnormalities in 60 to 85 percent of ALL cases, although
under optimal conditions the detection rate may approach 100
percent.
Some karyotypic changes are random or have no known prognostic
significance. Others are recurrent and provide information about
both underlying leukemogenic mechanisms and clinical prognosis. Cytogenetic changes occurring in ALL include both abnormalities
in chromosome number, such as trisomies and monosomies, and
abnormalities in structure, such as deletions, translocations
and inversions.
Deletions commonly lead to loss of a tumor
suppressor. Translocations and inversions generally cause two
types of events. A proto-oncogene may be brought into proximity
with a TCR or immunoglobulin locus causing overexpression of
the intact gene; or the genes at the breakpoints of the
rearranged chromosomes, often transcription factors, may fuse to
form a new, chimeric protein that is oncogenic due to altered
properties and/or expression patterns.
Abnormalities of chromosome number in ALL (ploidy)
Ploidy can be assessed either by karyotype analysis as a count
of chromosome number or by flow cytometry using a measure
called the DNA index (DI), which is the ratio of fluorescence in
leukemic blasts compared to normal diploid cells in the G0/G1
phase. Normal diploid cells have 46 chromosomes and a DI of 1.0,
hyperdiploid cells have greater than 46 chromosomes and DI
greater than 1.0, and hypodiploid less than 46 chromosomes and DI <1.0. Hyperdiploid samples
are further classified as “low” and “high” hyperdiploid,
indicating 47 to 50 and greater than 50 chromosomes, respectively.
Hypodiploid cases constitute approximately 6
percent of pediatric.
Approximately 80 percent of pediatric cases lack only a single
chromosome, and their prognosis is similar to diploid cases.
However, those with less than 45 chromosomes have significantly worse
outcome, with the worst outcome in near-haploid cases (24 to 28
chromosomes). Rare cases with near triploidy (68 to 80 chromosomes)
or near tetraploidy (greater than 80 chromosomes) also have generally been
associated with very poor outcome.
Hyperdiploidy occurs in about 35 percent of pediatric ALL. It usually
coincides with other favorable risk factors, but retains
independent positive predictive value. Outcome is best for
“high” hyperdiploidy, which is variably defined in different
studies as more than between 50 to 55. Recently, a meta-analysis
proposed that favorable prognosis is attributable to particular
chromosome trisomies rather than total number, with simultaneous
occurrence of trisomies 4, 10 and 17 being most favorable.
Abnormalities of chromosome structure in ALL
TEL-AML1, t(12;21)(p13;q22)
The TEL-AML1 fusion protein formed by the t(12;21)(p13;q22)
translocation occurs in approximately 25 percent of childhood ALL,
making it the most frequent abnormality in childhood cancer.
Despite being the most frequent translocation in ALL, the
t(12;21) translocation was not identified until 1995 because, in
nearly all cases, the translocation is cryptic, involving a
genetic region too small to be detected by karyotype.
The TEL-AML1 fusion protein is widely expressed due to the TEL
promoter, and converts AML1 from a transcriptional activator to
a repressor. The protein appears to act in a dominant-negative
fashion, repressing transcriptional activation by the remaining
normal copy of AML1. The manner in which this mediates
leukemogenesis remains unclear. The low penetrance, prolonged
latency, and variable results of TEL-AML1 overexpression in
different mouse models suggest that it is an important
initiating step, but not sufficient for leukemic transformation.
Clinical significance of TEL-AML1
The prognostic significance of TEL-AML1 positivity has
been a subject of controversy. It initially was thought to be a
favorable prognostic subgroup, with reported survival exceeding
90 percent. Subsequent reports raised concern about the frequency of
late relapse, suggesting that survival might be equivalent or
inferior provided sufficiently long follow-up time. More
recently, it has been suggested that differences in outcomes are
attributable to variable treatment type and intensity with
better outcome occurring in regimens with significant use of L-asparaginase
and methotrexate. A recent prospective study using an intensive
treatment regimen found that TEL-AML1 was associated with
superior prognosis (5-year overall survival of 97 percent), but did not
retain independent prognostic significance after age and
presenting WBC were taken into account.
Another hallmark of TEL-AML1+ ALL is the tendency to relapse
late, and for the relapsed leukemia to demonstrate excellent
chemosensitivity and salvage rate. This pattern of late relapse
and good salvage has prompted the idea that the original
leukemia actually may be eradicated, and that the relapse in
fact represents evolution of a new leukemic clone from the pre-leukemic
TEL-AML1+ cell of origin. Detailed molecular mapping of paired
diagnostic and relapsed TEL-AML1+ ALL clones has supported this
hypothesis, demonstrating identical TEL-AML1 sequences but
distinct secondary mutations.
E2A-PBX1, t(1;19)(q23;p13)
The E2A-PBX1 fusion protein, associated with the
t(1;19)(q23;p13) translocation, is the second most common
translocation in pediatric ALL, occurring in approximately 6
percent of
all pre-B ALL. The fusion protein combines the two activation
domains of the transcription factor E2A on chromosome 19 with
the homeobox gene PBX1 on chromosome 1, resulting in a chimeric
transcription factor which strongly activates a subset of
homeobox genes normally regulated by PBX1.
Clinical significance of E2A-PBX1
The E2A-PBX1 fusion protein tends to be associated with other
known high risk factors, but in early studies it was found to
have independent adverse prognostic impact. On modern intensive
treatment regimens, however, survival is equivalent with cure
rates up to 90%.
BCR-ABL, t(9;22)(q34;q11)
The t(9;22) translocation was the first recurrent chromosomal
abnormality identified in human cancer, in 1960, in association
with chronic myelocytic leukemia (CML). This translocation,
known as the Philadelphia chromosome (Ph), is an essential
criterion in diagnosing CML. It also occurs in about 3
percent of
pediatric ALL. Ph is a significant adverse prognostic marker
with significantly lower induction rates, more frequent and
earlier relapse, and poorer overall survival.
The Philadelphia chromosome is formed by the in-frame fusion of
the 5’ portion of BCR (breakpoint cluster region) on
chromosome 22 to the 3’ portion of the tyrosine kinase C-ABL on
chromosome 9, a proto-oncogene which is part of the RAS
signaling pathway. The resulting BCR-ABL fusion protein causes
upregulation of ABL tyrosine kinase activity. Two main chimeric
BCR-ABL proteins occur in ALL, which differ in BCR breakpoint.
Breaks within the 5.8 kb major breakpoint cluster region (M-BCR),
occurring in CML and 25 percent of adult Ph+ ALL, form a 210 kD protein
known as p210. In the remainder of adult ALL and the majority of
pediatric ALL, the breakpoint occurs further upstream in the BCR
gene, in the minor breakpoint cluster region (m-BCR), forming a
185-190 kD protein known as p185 or more often p190.
Clinical significance of BCR-ABL
Allogeneic stem cell transplant generally has been
regarded as the only curative therapy in Ph+ ALL, and
generally is recommended in first complete remission (CR). If a
related donor is unavailable, an unrelated donor may be
considered. Molecular monitoring is vital during therapy, as the
fusion protein should not persist in remission ALL, unlike in CML. Molecular remission is significantly more likely to be
durable than cytogenetic remission with molecular positivity,
both pre- and post-transplant.
Treatment of CML and Ph+ ALL was revolutionized in 2001 by the advent of imatinib mesylate, also known as STI-571 or Gleevec. Imatinib, a selective tyrosine kinase inhibitor, was the first
molecularly targeted therapy to attain large-scale clinical
success, fulfilling the goals of antitumor selectivity and low
systemic toxicity. Despite its success, it has not been
effective as a single agent due to the rapid development of
resistance, and allogeneic stem cell transplant in first CR
remains the optimal curative therapy. Ongoing research is needed
to determine how best to integrate imatinib into chemotherapy
regimens, and whether transplant is necessary for patients with
a rapid and sustained molecular response.
MLL, 11q23 rearrangements
MLL gene rearrangements occur in 8 percent of pediatric ALL, and
constitute the most frequent abnormality in infant ALL,
occurring in 60 percent to 70 percent of cases. They also are associated with AML,
particularly secondary malignancies following anthracycline and
epipodophyllotoxin therapy. MLL-rearranged leukemias are unusual
in two respects: (1) the N-terminal of MLL forms a fusion
protein with the C-terminal of over 40 different partners,
including itself; and (2) MLL rearrangements are found in both
ALL and AML, whereas most other translocations are lineage
specific.
The MLL-AF4 fusion protein formed
by t(4;11) is the most common MLL translocation in ALL, making
up 70 percent of cases. Infant ALL with MLL rearrangement tends to be
associated with age less than six months, more often female
gender, massive tumor burden, organomegaly, frequent CNS
involvement, coexpression of myeloid antigens and CD10-negative
pro-B immunophenotype. MLL rearrangement is a poor prognostic
feature in pediatric ALL, including in children over one year of
age, although outcomes appear to be particularly poor for the
infant age group and for the t(4;11) translocation.
T ALL
T ALL makes up 12 percent of pediatric ALL. It occurs most often in
adolescents and young adults, frequently presenting with an
extremely high WBC, CNS involvement, a large mediastinal mass,
and marked lymphadenopathy and hepatosplenomegaly. Historically,
survival was dismal compared to B-lineage ALL. With
intensified therapies it has improved to approximately 70
percent to 75 percent.
However, traditional risk factors such as age and WBC used for
stratification in B-lineage ALL appear not to be as
prognostically informative in T ALL, highlighting the importance
of identifying molecularly based prognostic differences instead.
The leukemogenic event in T ALL
typically involves overexpression of an unaltered
proto-oncogene, rather than generation of a novel fusion protein
as frequently occurs in B-lineage ALL. Often, overexpression
occurs due to a translocation placing a gene under the control
of a TCR promoter or enhancer. The two TCR loci most frequently
involved are the TCRb-chain locus at 7q34 and the TCRa/d-chain
locus at 14q11.
Conclusions
Unraveling the molecular genetics of ALL has paved the way
toward many advances in our understanding of basic pathways of
leukemogenesis; our ability to stratify patients at diagnosis
into appropriately tailored treatment groups and track disease
status during treatment; and the identification of novel
therapeutic targets. Future challenges include searching for new
insights into optimizing therapy for existing poor-risk disease
subtypes as well as identifying patients within favorable-risk
subtypes who nevertheless fail to be cured, and devising more
effective therapies for them. On a practical level, it will be
an increasing challenge to take the mushrooming quantity of
molecular and genomic research on ALL and extract those
principles which are most informative and feasible for
large-scale application in the clinical realm.
About the authors
Karen Rabin, M.D., and
Judith Margolin, M.D., are assistant
professors of pediatric hematology/oncology and members of the
Texas Children's Cancer Center's
Leukemia and Lymphoma Team.
Dr. Rabin's clinical interest is ALL. Her research
with Down
syndrome focuses on
the molecular genetics of ALL in patients
and on clinical applications of array comparative genomic
hybridization.
Dr. Margolin is interested in
the patterns of gene expression, which occur in normal
hemopoietic cells during development, differentiation and immune
response. Her lab also studies gene expression patterning in
leukemic blast cells, hepatoblastoma cells and other pediatric
cancers.
Back to top |
