TREATMENT READING
Childhood Acute Lymphoblastic Leukemia
Source sections for this topic, with the original link and date. Patient and professional reading are labelled separately.
Not medical advice · Not a complete treatment or trial directory
Made by Triangle.
Selected text from the National Cancer Institute, with the original link and date. Made by Triangle. This page is not an NCI PDQ summary. Source topics may be broader than an individual diagnosis.
153 sections
- Risk-Based Treatment Assignment
- Introduction to Risk-Based Treatment
- Prognostic Factors Affecting Risk-Based Treatment
- Patient and clinical disease characteristics
- Age at diagnosis
- WBC count at diagnosis
- CNS involvement at diagnosis
- Testicular involvement at diagnosis
- Down syndrome (trisomy 21)
- Sex
- Race and ethnicity
- Weight at diagnosis and during treatment
- Leukemic characteristics
- Immunophenotype
- Cytogenetics/genomic alterations
- Response to initial treatment
- MRD determination in bone marrow at the end of induction (EOI) and end of consolidation (EOC)
- Day 7 and day 14 bone marrow responses
- Peripheral blood response to steroid prophase
- Peripheral blood response to multiagent induction therapy
- Peripheral blood MRD before EOI (day 8, day 15)
- Persistent leukemia at the EOI (induction failure)
- Flow cytometry versus morphology
- Prognostic (Risk) Groups
- Children’s Oncology Group (COG) risk groups
- Berlin-Frankfurt-Münster (BFM) risk groups
- Prognostic (risk) groups under clinical evaluation
- Treatment Option Overview for Childhood ALL
- Phases of Therapy
- Sanctuary Sites
- Central nervous system (CNS)
- Testes
- Special Considerations for the Treatment of Children With ALL
- Treatment of Newly Diagnosed Childhood ALL
- Standard Induction Treatment Options for Newly Diagnosed ALL
- Remission induction chemotherapy
- Corticosteroid therapy
- Asparaginase
- Pegaspargase (PEG-asparaginase)
- Calaspargase pegol
- Asparaginase Erwinia chrysanthemi (Erwinia L-asparaginase)
- Anthracycline use during induction
- Response to remission induction chemotherapy
- Standard Postinduction Treatment Options for Childhood ALL
- Consolidation/intensification therapy
- Standard-risk ALL
- High-risk ALL
- Very high-risk ALL
- Maintenance therapy
- Backbone of maintenance therapy
- Vincristine/corticosteroid pulses
- Duration of maintenance therapy
- Adherence to oral medications during maintenance therapy
- Allopurinol for mercaptopurine (6-MP) toxicity
- Treatment options under clinical evaluation
- Central Nervous System (CNS)-Directed Therapy for Childhood ALL
- Overview of CNS-Directed Treatment Regimens
- Intrathecal Chemotherapy
- CNS-Directed Systemic Chemotherapy
- Cranial Radiation Therapy
- CNS Therapy for Standard-Risk Patients
- CNS Therapy for High-Risk and Very High-Risk Patients Without CNS Involvement
- Intrathecal chemotherapy
- Cranial radiation therapy
- CNS Therapy for Patients With CNS3 Disease at Diagnosis
- CNS Therapy Options Under Clinical Evaluation
- Toxicity of CNS-Directed Therapy
- Acute and subacute toxicities
- Late-developing toxicities
- Postinduction Treatment for Specific ALL Subgroups
- T-ALL
- Treatment options for T-ALL
- Treatment options under clinical evaluation for T-ALL
- Infants With ALL
- Treatment options for infants with KMT2A rearrangements
- Treatment options for infants without KMT2A rearrangements
- Treatment options under clinical evaluation for infants with ALL
- Adolescents and Young Adults With ALL
- Treatment options for adolescents and young adults with ALL
- Osteonecrosis
- Treatment options under clinical evaluation for adolescent and young adult patients with ALL
- Children With Down Syndrome
- Treatment of children with Down syndrome
- Treatment options under clinical evaluation for children with Down syndrome and ALL
- BCR::ABL1-Positive (Philadelphia Chromosome–Positive) ALL
- Treatment options for patients with BCR::ABL1-positive ALL
- Treatment options under clinical evaluation for BCR::ABL1 ALL
- Treatment of Relapsed Childhood ALL
- Prognostic Factors After First Relapse of Childhood ALL
- Site of relapse
- Time from diagnosis to relapse
- Patient characteristics
- Risk group classification at initial diagnosis
- Response to reinduction therapy
- Cytogenetics/genomic alterations
- Immunophenotype
- Standard Treatment Options for First Bone Marrow Relapse of Childhood ALL
- Reinduction chemotherapy
- Postreinduction therapy for patients achieving a second complete remission
- Early-relapsing B-ALL
- Late-relapsing B-ALL
- BCR::ABL1 (Philadelphia chromosome–positive [Ph+]) ALL
- T-ALL
- Treatment Options for Second and Subsequent Bone Marrow Relapse
- Hematopoietic Stem Cell Transplant for First and Subsequent Bone Marrow Relapse
- Components of the transplant process
- TBI-containing transplant preparative regimens
- MRD detection just before transplant
- MRD detection posttransplant
- Donor type and HLA match
- Role of GVHD/GVL in ALL and immune modulation after transplant to prevent relapse
- Intrathecal medication after HSCT to prevent relapse
- Relapse after allogeneic HSCT for relapsed ALL
- Immunotherapeutic Approaches for Relapsed or Refractory ALL
- Monoclonal antibody therapy
- CAR T-cell therapy
- Toxicities associated with CAR T-cell therapy
- CD19-targeted CAR T-cell therapy
- CD19-targeted CAR T-cell treatment for CNS and other extramedullary disease
- Risk factors associated with CD19-targeted CAR T-cell failure
- Relapse or failure to respond after CD19-targeted CAR T-cell therapy
- Prevention of relapse or failure after CD19-targeted CAR T-cell therapy
- Previous allogeneic donor–derived CD19-targeted CAR T-cell therapy
- The role of consolidative HSCT after CD19-targeted CAR T-cell therapy for ALL
- CD22-targeted CAR T-cell therapy
- Multiantigen targeted CAR T-cell therapy
- Treatment of Isolated Extramedullary Relapse
- Isolated CNS relapse
- Isolated testicular relapse
- Treatment Options Under Clinical Evaluation for Relapsed Childhood ALL
- Trials for ALL in first relapse
- Trials for ALL in second or subsequent relapse or refractory ALL
- Types of treatment for childhood acute lymphoblastic leukemia
- Who treats children with acute lymphoblastic leukemia?
- Treatment phases of childhood acute lymphoblastic leukemia
- Chemotherapy
- Radiation therapy
- Stem cell transplant
- Immunotherapy
- Targeted therapy
- Clinical trials
- Treatment options for childhood acute lymphoblastic leukemia
- Treatment of standard-risk childhood acute lymphoblastic leukemia
- Treatment of high-risk childhood acute lymphoblastic leukemia
- Treatment of very high-risk childhood acute lymphoblastic leukemia
- Treatment of childhood acute lymphoblastic leukemia in the brain and spinal cord or testicles
- Treatment of T-cell childhood acute lymphoblastic leukemia
- Treatment of infants with acute lymphoblastic leukemia
- Treatment of adolescents and young adults with acute lymphoblastic leukemia
- Treatment of children with Down syndrome and acute lymphoblastic leukemia
- Treatment of childhood Philadelphia chromosome–positive acute lymphoblastic leukemia
- Treatment of relapsed or refractory childhood acute lymphoblastic leukemia
- Side effects and late effects of treatment
Risk-Based Treatment Assignment
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Introduction to Risk-Based Treatment
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Children with acute lymphoblastic leukemia (ALL) are usually treated according to risk groups defined by both clinical and laboratory features. The intensity of treatment required for cure varies substantially among subsets of children with ALL. Risk-based treatment assignment is used in children with ALL so that patients with favorable clinical and biological features who are likely to have a very good outcome with modest therapy can be spared more intensive and toxic treatment, while a more aggressive, potentially more toxic therapeutic approach is reserved for patients with a lower probability of long-term survival.[1,2]
Certain ALL study groups, such as the Children’s Oncology Group (COG), use a more- or less-intensive induction regimen based on a subset of pretreatment factors, while other groups give a similar induction regimen to all patients.
Factors used by the COG to determine the intensity of induction include the following:
Immunophenotype.
The presence or absence of extramedullary disease.
Steroid pretreatment.
The presence or absence of Down syndrome.
The National Cancer Institute (NCI) risk group classification.
The NCI risk group classification for B-ALL stratifies risk according to age and white blood cell (WBC) count, as follows:[3]
Sources: 3
Standard risk: WBC count less than 50,000/μL and age 1 to younger than 10 years.
High risk: WBC count 50,000/μL or greater and/or age 10 years or older.
All study groups modify the intensity of postinduction therapy on the basis of a variety of prognostic factors, including NCI risk group, immunophenotype, early response determinations, and cytogenetics and genomic alterations.[4] Detection of the BCR::ABL1 fusion (i.e., BCR::ABL1-positive ALL) leads to immediate changes in induction therapy, including the addition of a tyrosine kinase inhibitor, such as imatinib or dasatinib.[5]
Risk-based treatment assignment requires the availability of prognostic factors that reliably predict outcome. For children with ALL, a number of factors have demonstrated prognostic value, some of which are described below.[6] Factors affecting prognosis are grouped into the following three categories:
Sources: 6
Patient and clinical disease characteristics.
Leukemic characteristics.
Sources: Leukemic characteristics.
Response to initial treatment.
Sources: Response to initial treatment.
As in any discussion of prognostic factors, the relative order of significance and the interrelationship of the variables are often treatment dependent and require multivariate analysis to determine which factors operate independently as prognostic variables. Because prognostic factors are treatment dependent, improvements in therapy may diminish or abrogate the significance of any of these prognostic factors.
A subset of the prognostic and clinical factors discussed below is used for the initial stratification of children with ALL for treatment assignment. For brief descriptions of the prognostic groupings currently applied in ongoing clinical trials in the United States, see the Prognostic (risk) groups under clinical evaluation section.
For information about important prognostic factors at relapse, see the Prognostic Factors After First Relapse of Childhood ALL section.
Sources: Prognostic Factors After First Relapse of Childhood ALL
This page is made by Triangle. The reading above is selected source text, not a treatment recommendation, and it is not medical advice.
Prognostic Factors Affecting Risk-Based Treatment
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Patient and clinical disease characteristics
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Patient and clinical disease characteristics affecting prognosis include the following:
Age at diagnosis.
Sources: Age at diagnosis
WBC count at diagnosis.
Sources: WBC count at diagnosis
Central nervous system (CNS) involvement at diagnosis.
Sources: Central nervous system (CNS) involvement at diagnosis
Testicular involvement at diagnosis.
Sources: Testicular involvement at diagnosis
Down syndrome (trisomy 21).
Sources: Down syndrome (trisomy 21)
Sex.
Sources: Sex
Race and ethnicity.
Sources: Race and ethnicity
Weight at diagnosis and during treatment.
Age at diagnosis
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Age at diagnosis has strong prognostic significance in patients with B-ALL, reflecting the different underlying biology of ALL in different age groups.[7] Age at diagnosis is not prognostically relevant in T-ALL.[8]
Infants (younger than 1 year). Infants with ALL have a particularly high risk of treatment failure. Treatment failure is most common in the following groups: Up to 80% of infants with ALL have a translocation of 11q23 with numerous chromosome partners generating a KMT2A gene rearrangement.[10,12,14,15] The most common rearrangement is KMT2A::AFF1 (t(4;11)(q21;q23)), but KMT2A rearrangements with many other translocation partners are observed. Infants with leukemia and KMT2A rearrangements typically have very high WBC counts and an increased incidence of CNS involvement. Event-free survival (EFS) and overall survival (OS) rates are poor. The 5-year EFS and OS rates are 35% to 40% for infants with KMT2A-rearranged ALL.[10-12] The frequency of KMT2A gene rearrangements is extremely high in infants younger than 6 months. From 6 months to 1 year, the incidence of KMT2A rearrangements decreases but remains significantly higher than that observed in older children.[10,16] Blasts from infants with KMT2A rearrangements are often CD10 negative and express high levels of FLT3.[10,11,15,17] Conversely, infants whose leukemic cells show a germline KMT2A gene configuration frequently present with CD10-positive precursor-B immunophenotype. These infants have a significantly better outcome than do infants with ALL characterized by KMT2A rearrangements.[10,11,15,18] Black infants with ALL are significantly less likely to have KMT2A rearrangements than White infants.[16] A comparison of the landscape of somatic variants in infants and older children with KMT2A-rearranged ALL revealed significant differences between the two groups. This result suggests distinctive age-related biological behaviors for KMT2A-rearranged ALL that may relate to the significantly poorer outcome for infants.[19,20] For more information about infants with ALL, see the Infants With ALL section.
Sources: 10 · 12 · 14 · 15 · 10 · 12 · 10 · 16 · 10 · 11 · 15 · 17 · 10 · 11 · 15 · 18 · 16 · 19 · 20 · Infants With ALL
Infants (younger than 1 year). Infants with ALL have a particularly high risk of treatment failure. Treatment failure is most common in the following groups: Up to 80% of infants with ALL have a translocation of 11q23 with numerous chromosome partners generating a KMT2A gene rearrangement.[10,12,14,15] The most common rearrangement is KMT2A::AFF1 (t(4;11)(q21;q23)), but KMT2A rearrangements with many other translocation partners are observed. Infants with leukemia and KMT2A rearrangements typically have very high WBC counts and an increased incidence of CNS involvement. Event-free survival (EFS) and overall survival (OS) rates are poor. The 5-year EFS and OS rates are 35% to 40% for infants with KMT2A-rearranged ALL.[10-12] The frequency of KMT2A gene rearrangements is extremely high in infants younger than 6 months. From 6 months to 1 year, the incidence of KMT2A rearrangements decreases but remains significantly higher than that observed in older children.[10,16] Blasts from infants with KMT2A rearrangements are often CD10 negative and express high levels of FLT3.[10,11,15,17] Conversely, infants whose leukemic cells show a germline KMT2A gene configuration frequently present with CD10-positive precursor-B immunophenotype. These infants have a significantly better outcome than do infants with ALL characterized by KMT2A rearrangements.[10,11,15,18] Black infants with ALL are significantly less likely to have KMT2A rearrangements than White infants.[16] A comparison of the landscape of somatic variants in infants and older children with KMT2A-rearranged ALL revealed significant differences between the two groups. This result suggests distinctive age-related biological behaviors for KMT2A-rearranged ALL that may relate to the significantly poorer outcome for infants.[19,20] For more information about infants with ALL, see the Infants With ALL section.
Infants younger than 6 months (with an even poorer prognosis for those aged ≤90 days).[9-13]
Infants with extremely high presenting leukocyte counts (>200,000–300,000 × 109/L).[10]
Sources: 10
Infants with a poor response to a prednisone prophase.[10]
Sources: 10
Young children (aged 1 to <10 years). Young children (aged 1 to <10 years) with B-ALL have a better disease-free survival (DFS) rate than older children, adolescents, and infants.[3,7,21-23] The improved prognosis in young children is at least partly explained by the more frequent occurrence of favorable cytogenetic features in the leukemic blasts, including hyperdiploidy with 51 to 65 chromosomes and/or favorable chromosome trisomies, or the ETV6::RUNX1 fusion (t(12;21)(p13;q22), previously known as the TEL::AML1 translocation).[7,24,25]
Adolescents and young adults (aged ≥10 years). In general, the outcome of patients with B-ALL aged 10 years and older is inferior to that of patients aged 1 to younger than 10 years.[26] Patients aged 10 to 15 years fare better than those who are aged 16 to 21 years at diagnosis who were treated with pediatric regimens.[8] However, the outcome for older adolescents has improved significantly over time.[27-29] Five-year survival rates for adolescents aged 15 to 19 years increased from 36% (1975–1984) to 78% (2011–2017).[30-33] Multiple retrospective studies have established that adolescents aged 16 to 21 years have a better outcome when treated on pediatric versus adult protocols.[34-36] For more information about adolescents with ALL, see the Postinduction Treatment for Specific ALL Subgroups section.
Sources: 26 · 8 · 27 · 29 · 30 · 33 · 34 · 36 · Postinduction Treatment for Specific ALL Subgroups
WBC count at diagnosis
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
A WBC count of 50,000/µL is generally used as an operational cut point between better and poorer prognosis,[3] although the relationship between WBC count and prognosis is a continuous function rather than a step function. Patients with B-ALL and high WBC counts at diagnosis have an increased risk of treatment failure compared with patients with low initial WBC counts.[37]
CNS involvement at diagnosis
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The presence or absence of CNS leukemia at diagnosis has prognostic significance in both patients with B-ALL and T-ALL. Patients who have a nontraumatic diagnostic lumbar puncture may be placed into one of three categories according to the number of WBC/µL and the presence or absence of blasts on cytospin as follows:
CNS1: Cerebrospinal fluid (CSF) that is cytospin negative for blasts regardless of WBC count.
CNS2: CSF with fewer than 5 WBC/µL and cytospin positive for blasts.
CNS3 (CNS disease): CSF with 5 or more WBC/µL and cytospin positive for blasts or clinical signs of CNS leukemia (i.e., facial nerve palsy, brain/eye involvement, or hypothalamic syndrome).
Children with B-ALL or T-ALL who present with CNS3 disease at diagnosis are at a higher risk of treatment failure (both within the CNS and systemically) than patients who are classified as CNS1 or CNS2.[47-49] The prognostic implication of CNS2 status at diagnosis may differ between patients with B-ALL and T-ALL. Some studies have reported increased risk of CNS relapse and/or inferior EFS in patients with B-ALL and CNS2 status at diagnosis, compared with patients with CNS1 status,[50,51] while other studies have not.[47,52-54] In an analysis of 2,164 patients with T-ALL treated in two consecutive COG trials, there was no difference in EFS, DFS, or cumulative incidence of relapse between patients with CNS1 and CNS2 status.[49]
A traumatic lumbar puncture (≥10 erythrocytes/µL) that includes blasts at diagnosis has also been associated with increased risk of CNS relapse and overall poorer outcome in some studies,[47,53,55] but not others.[51,52,56] Patients with CNS2, CNS3, or traumatic lumbar puncture have a higher frequency of unfavorable prognostic characteristics than those with CNS1, including significantly higher WBC counts at diagnosis, older age at diagnosis, an increased frequency of the T-ALL phenotype, and KMT2A gene rearrangements.[47,52,53]
Most clinical trial groups have approached the treatment of CNS2 and traumatic lumbar puncture patients by using more intensive therapy, primarily additional doses of intrathecal therapy during induction.[47,57,58]; [52][Level of evidence B4]; [59][Level of evidence A1]
Sources: 47 · 57 · 58 · 52 · Level of evidence B4 · 59 · Level of evidence A1
To determine whether a patient with a traumatic lumbar puncture (with blasts) should be treated as CNS3, the COG uses an algorithm relating the WBC and red blood cell counts in the spinal fluid and the peripheral blood.[60]
Sources: 60
Testicular involvement at diagnosis
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Overt testicular involvement at the time of diagnosis occurs in approximately 2% of males,[61,62] with a higher frequency in patients with T-ALL than in patients with B-ALL.[62]
In early ALL trials, testicular involvement at diagnosis was an adverse prognostic factor. With more aggressive initial therapy, however, it does not appear to have prognostic significance.[61,62] For example, a European Organization for Research and Treatment of Cancer trial (EORTC-58881) reported no adverse prognostic significance for overt testicular involvement at diagnosis.[62]
The role of radiation therapy for testicular involvement is unclear. A study from St. Jude Children's Research Hospital (SJCRH) suggests that a good outcome can be achieved with aggressive conventional chemotherapy without radiation.[61] The COG has also adopted this strategy for boys with testicular involvement that resolves completely by the end of induction therapy. The COG considers patients with testicular involvement to be high risk regardless of other presenting features, but most other large clinical trial groups in the United States and Europe do not consider testicular disease to be a high-risk feature.
Sources: 61
Down syndrome (trisomy 21)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Outcomes in children with Down syndrome and ALL have often been somewhat inferior to outcomes in children without Down syndrome.[63-67] However, in some studies, patients with Down syndrome appeared to fare as well as those without Down syndrome.[68,69] The lower EFS and OS of children with Down syndrome appear to be related to increased frequency of treatment-related mortality, as well as higher rates of induction failure and relapse.[63-66,70,71] The inferior anti-leukemic outcome may be due, in part, to the decreased prevalence of favorable biological features such as ETV6::RUNX1 or hyperdiploidy (51–65 chromosomes) with trisomies of chromosomes 4 and 10 in Down syndrome ALL patients.[70-72]
In a large retrospective study that included 653 patients with Down syndrome and ALL, patients with Down syndrome had a lower complete remission (CR) rate (97% vs. 99%, P < .001), higher cumulative incidence of relapse (26% vs. 15%, P < .001) and higher treatment-related mortality (7% vs. < 1%, P < .001) compared with patients without Down syndrome.[71] Among the patients with Down syndrome, age younger than 6 years, WBC count of less than 10,000/µL, and the presence of the ETV6::RUNX1 fusion (observed in 8% of patients) were independent predictors of favorable EFS.
Sources: 71
In a report from the COG, among patients with B-ALL who lacked KMT2A rearrangements, BCR::ABL1, ETV6::RUNX1, and hyperdiploidy with trisomies of chromosomes 4 and 10, the EFS and OS rates were similar in children with and without Down syndrome.[70]
Sources: 70
Certain genomic abnormalities, such as IKZF1 deletions, CRLF2 aberrations, and JAK variants, are seen more frequently in ALL arising in children with Down syndrome than in those without Down syndrome.[73-77] Studies of children with Down syndrome and ALL suggest that the presence of IKZF1 deletions (but not CRLF2 aberrations or JAK variants) is associated with an inferior prognosis.[71,77,78]
A retrospective analysis included 130 patients with CRLF2-rearranged ALL and Down syndrome. Patients with the BCR::ABL1-like signature (25% of the CRLF2-rearranged cases) had an inferior outcome compared with those who lacked the BCR::ABL1-like signature (EFS rates, 39.5% ± 8.1% vs. 82.0% ± 4.4%; hazard ratio [HR], 5.27).[72]
Sources: 72
The IGH::IGF2BP1 gene fusion occurs in approximately 3% of patients with Down syndrome. This fusion is rare in patients with ALL who do not have Down syndrome. In one retrospective analysis, the fusion was associated with a relatively favorable outcome (EFS rate, 87.5% ± 1.7%).[72]
Sources: 72
Sex
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
In some studies, the prognosis for girls with ALL is slightly better than it is for boys with ALL.[79-81] One reason is the occurrence of testicular relapses among boys, but boys also appear to be at increased risk of bone marrow and CNS relapse for reasons that are not well understood.[79-81] While some reports describe outcomes for boys as closely approaching those of girls,[23,57,82] larger clinical trial experiences and national data continue to show somewhat lower survival rates for boys.[22,33,83,84]
Sources: 79 · 81 · 79 · 81 · 23 · 57 · 82 · 22 · 33 · 83 · 84
Race and ethnicity
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Over the last several decades in the United States, survival rates in Black and Hispanic children with ALL have been somewhat lower than those in White children with ALL.[85-88] One study included more than 18,031 patients with B-ALL and 1,892 patients with T-ALL who were aged 0 to 30 years and treated between 2004 and 2019 in COG clinical trials. The race- and ethnicity-based outcome disparities noted in older studies persisted with more contemporary therapy. Race- and ethnicity-based outcome disparities were observed for patients with B-ALL but not for patients with T-ALL. The study also noted a wider disparity in OS versus EFS for patients with B-ALL, suggesting that disparities might be greater in the setting of relapsed disease versus newly diagnosed disease.[89] Multivariable analysis adjusting for disease prognosticators (e.g., age and WBC count, cytogenetic risk group, CNS status) and insurance status substantially attenuated the increased risk of inferior EFS for Hispanic patients. However, the same adjustments did not attenuate the inferior EFS for non-Hispanic Black children.[89]
The following factors associated with race and ethnicity influence survival:
ALL subtype. The reason for better outcomes in White and Asian children than in Black and Hispanic children is at least partially explained by the different spectrum of ALL subtypes. For example, Black children have a higher relative incidence of T-ALL, lower rates of favorable genetic subtypes of B-ALL, and higher rates of the TFC3::PBX1 (t(1;19)) translocation. Hispanic and Latino children have a lower prevalence of the favorable ETV6::RUNX1 fusion gene.[90] Hispanic and Latino children have a higher frequency of CRLF2 rearrangements and IKZF1 deletions.[90-92] They also have a higher frequency of the IKZF1PLUS profile (IKZF1 deletion plus deletion of CDKN2A, CDKN2B, PAX5, or PAR1 [in the absence of ERG deletion]).[90]
Treatment adherence. Differences in outcome may also be related to treatment adherence, as illustrated by a study of adherence to oral mercaptopurine (6-MP) in maintenance therapy. In the first report from the study, there was an increased risk of relapse in Hispanic children compared with non-Hispanic White children, depending on the level of adherence, even when adjusting for other known variables. However, even with adherence rates of 90% or more, Hispanic children continued to demonstrate increased rates of relapse.[93] In the second report from the study, adherence rates were shown to be significantly lower in Asian American and African American patients than in non-Hispanic White patients. A greater percentage of patients in these ethnic groups had adherence rates of less than 90%, which was associated with a 3.9-fold increased risk of relapse.[94]
Ancestry-related genomic variations. Ancestry-related genomic variations may also contribute to racial and ethnic disparities in both the incidence and outcome of ALL.[95] For example, the differential presence of specific host polymorphisms in different racial and ethnic groups may contribute to outcome disparities, as illustrated by the occurrence of single nucleotide polymorphisms in the ARID5B gene that occur more frequently among Hispanic patients and are linked to both ALL susceptibility and to relapse hazard.[96] In a genome-wide association study (GWAS), the GATA3 variant, rs3824662, was associated with an increased risk of developing BCR::ABL1-like (Ph-like) ALL. Patients with this variant were at increased risk of high minimal residual disease (MRD) at end-induction and at greater risk of relapse. The rs3824662 risk allele is associated with Native American genetic ancestry. The risk allele frequency was 52% in Guatemalan patients, 40% in U.S. Hispanic patients, and 14% in patients of European descent.[97]
Weight at diagnosis and during treatment
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Studies of the impact of obesity on the outcome of ALL have had variable results. In most of these studies, obesity is defined as weight above the 95th percentile for age and height.
Three studies did not demonstrate an independent effect of obesity on EFS.[98][Level of evidence B4]; [99,100][Level of evidence C2]
Sources: 98 · Level of evidence B4 · 99 · 100 · Level of evidence C2
Two studies showed obesity to be an independent prognostic factor only in patients older than 10 years or in patients with intermediate-risk or high-risk disease.[101,102][Level of evidence C2]
Sources: 101 · 102 · Level of evidence C2
The COG reported on the impact of obesity on outcome in 2,008 children, 14% of whom were obese, who were enrolled on a high-risk ALL trial (CCG-1961 [NCT00002812]).[103][Level of evidence B4] Obesity was found to be an independent variable for inferior outcome compared with nonobese patients (5-year EFS rates, 64% vs. 74%; P = .002). However, obese patients at diagnosis who then lost weight during the premaintenance period of treatment had outcomes similar to patients with normal weight at diagnosis.
Sources: CCG-1961 [NCT00002812] · 103 · Level of evidence B4
In a retrospective study of patients treated at a single institution, obesity at diagnosis was linked to an increased risk of having MRD at the end of induction and an inferior EFS.[104][Level of evidence C2]
Sources: 104 · Level of evidence C2
In a different retrospective study of 373 patients treated at a single institution, body mass index (BMI) at diagnosis was not associated with MRD at days 19 and 46, cumulative incidence of relapse, or EFS. OS was lower in patients with a high BMI, primarily resulting from treatment-related mortality and inferior salvage after relapse.[105][Level of evidence C1]
Sources: 105 · Level of evidence C1
In one study, obesity at diagnosis was associated with increased toxicity and truncated administration of asparaginase, especially in older children and adolescents.[106]
Sources: 106
In a study of 388 patients aged 15 to 50 years who were treated with Dana-Farber Cancer Institute (DFCI) ALL consortium regimens, greater BMI was associated with higher rates of relapse and nonrelapse mortality, as well as inferior OS. Higher BMI was associated with increased rates of hepatotoxicity and hyperglycemia. The deleterious effect of elevated BMI was more pronounced in older patients. Among patients aged 15 to 29 years at diagnosis (n = 254), the 4-year OS rate was 73% for those with high BMI, compared with 83% for those with BMI in the reference range (P = .09).[107]
Sources: 107
In a study of 762 pediatric patients with ALL (aged 2–17 years), the Dutch Childhood Oncology Group found that those who were underweight at diagnosis (defined as BMI standard deviation score < -1.8; 8% of the population) had an almost twofold higher risk of relapse compared with patients who were not underweight (after adjusting for risk group and age), although this did not result in a difference in EFS or OS. Patients with a decrease in BMI during the first 32 weeks of treatment had similar rates of relapse as other patients, but had significantly worse OS, primarily because of poorer salvage rates after relapse.[108]
Sources: 108
Leukemic characteristics
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Leukemic cell characteristics affecting prognosis include the following:
Immunophenotype.
Sources: Immunophenotype
Cytogenetics/genomic alterations.
Sources: Cytogenetics/genomic alterations
Immunophenotype
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The 2016 revision to the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia classifies ALL as either B-lymphoblastic leukemia or T-lymphoblastic leukemia, with further subdivisions based on molecular characteristics.[109,110] For more information, see the Diagnosis section.
Either B- or T-lymphoblastic leukemia can coexpress myeloid antigens. These cases need to be distinguished from leukemia of ambiguous lineage.
B-ALL (WHO B-lymphoblastic leukemia). Before 2008, the WHO classified B-lymphoblastic leukemia as precursor B-lymphoblastic leukemia, and this terminology is still frequently used in the literature of childhood ALL to distinguish it from mature B-cell ALL. Mature B-cell ALL is now termed Burkitt leukemia and requires different treatment than has been given for B-ALL (precursor B-cell ALL). B-ALL, defined by the expression of CD19, HLA-DR, cytoplasmic CD79a, and other B-cell–associated antigens, accounts for 80% to 85% of childhood ALL. Approximately 90% of B-ALL cases express the CD10 surface antigen (formerly known as common ALL antigen). Absence of CD10 is often associated with KMT2A rearrangements, particularly t(4;11)(q21;q23), and a poor outcome.[10,111] It is not clear whether CD10 negativity has any independent prognostic significance in the absence of a KMT2A gene rearrangement.[112] The major immunophenotypic subtypes of B-ALL are as follows:
B-ALL (WHO B-lymphoblastic leukemia). Before 2008, the WHO classified B-lymphoblastic leukemia as precursor B-lymphoblastic leukemia, and this terminology is still frequently used in the literature of childhood ALL to distinguish it from mature B-cell ALL. Mature B-cell ALL is now termed Burkitt leukemia and requires different treatment than has been given for B-ALL (precursor B-cell ALL). B-ALL, defined by the expression of CD19, HLA-DR, cytoplasmic CD79a, and other B-cell–associated antigens, accounts for 80% to 85% of childhood ALL. Approximately 90% of B-ALL cases express the CD10 surface antigen (formerly known as common ALL antigen). Absence of CD10 is often associated with KMT2A rearrangements, particularly t(4;11)(q21;q23), and a poor outcome.[10,111] It is not clear whether CD10 negativity has any independent prognostic significance in the absence of a KMT2A gene rearrangement.[112] The major immunophenotypic subtypes of B-ALL are as follows:
Common B-ALL (CD10 positive and no surface or cytoplasmic immunoglobulin [Ig]).Approximately three-quarters of patients with B-ALL have the common precursor B-cell immunophenotype and have the best prognosis. Patients with favorable cytogenetics almost always show a common precursor B-cell immunophenotype.
Pro-B ALL (CD10 negative and no surface or cytoplasmic Ig).Approximately 5% of patients have the pro-B immunophenotype. Pro-B is the most common immunophenotype seen in infants and is often associated with KMT2A gene rearrangements.
Pre-B ALL (presence of cytoplasmic Ig).The leukemic cells of patients with pre-B ALL contain cytoplasmic Ig, and 25% of patients with pre-B ALL have the t(1;19)(q21;p13) translocation with the TCF3::PBX1 fusion.[113,114]Approximately 3% of patients have transitional pre-B ALL with expression of surface Ig heavy chain in the absence of Ig light chain expression, MYC gene involvement, and L3 morphology. Patients with this phenotype respond well to therapy used for B-ALL.[115]
Mature B-ALL (Burkitt lymphoma/leukemia).Approximately 2% of patients present with mature B-cell leukemia (surface Ig expression, generally with French-American-British criteria L3 morphology and an 8q24 translocation involving MYC), also called Burkitt leukemia. The treatment for mature B-cell ALL is based on therapy for non-Hodgkin lymphoma and is completely different from the treatment for B-ALL. Rare cases of mature B-cell leukemia that lack surface Ig but have L3 morphology with MYC gene translocations should also be treated as mature B-cell leukemia.[115] For more information about the treatment of children with mature B-cell lymphoma/leukemia and Burkitt lymphoma/leukemia, see Childhood Non-Hodgkin Lymphoma Treatment.A small number of cases of IG::MYC-translocated leukemias with precursor B-cell immunophenotype (e.g., absence of CD20 expression and surface Ig expression) have been reported.[116] These cases presented in both children and adults. Like Burkitt lymphoma/leukemia, they had a male predominance and most patients showed L3 morphology. The cases lacked variants in genes recurrently altered in Burkitt lymphoma (e.g., ID3, CCND3, or MYC), whereas variants in RAS genes (frequently altered in B-ALL) were common. The clinical significance and optimal therapy of IG::MYC–translocated leukemias with precursor B-cell phenotype and molecular characteristics requires further study.
Sources: 115 · Childhood Non-Hodgkin Lymphoma Treatment · 116
T-ALL. T-ALL is defined by expression of the T-cell–associated antigens (cytoplasmic CD3, with CD7 plus CD2 or CD5) on leukemic blasts. T-ALL is frequently associated with a constellation of clinical features, including the following:[21,39,82] While not true historically, with appropriately intensive therapy, children with T-ALL now have an outcome approaching that of children with B-ALL.[21,39,42,43,82,117] There are few commonly accepted prognostic factors for patients with T-ALL. Conflicting data exist regarding the prognostic significance of presenting leukocyte counts in T-ALL.[38-45,118] The presence or absence of a mediastinal mass at diagnosis has no prognostic significance. In patients with a mediastinal mass, the rate of regression of the mass lacks prognostic significance.[119] Early T-cell precursor (ETP) ALL. ETP ALL, a distinct subset of childhood T-ALL, was initially defined by identifying T-ALL cases with gene expression profiles highly related to expression profiles for normal early T-cell precursors.[120] The subset of T-ALL cases identified by these analyses represented 13% of all cases, and are characterized by a distinctive immunophenotype (CD1a and CD8 negativity, with weak expression of CD5 and coexpression of stem cell or myeloid markers). Another subgroup of T-ALL, called near-ETP ALL, has a similar immunophenotype as ETP ALL, except with strong CD5 expression. This subtype represents approximately 15% of cases.[46] Initial reports describing ETP ALL suggested that this subset of patients has a poorer prognosis than other patients with T-ALL.[120-122] In addition, studies have reported that patients with ETP and near-ETP ALL have a slower early response and higher frequency of induction failure.[45,46] However, despite higher rates of end-induction MRD and induction failure in these patients, the ETP and near-ETP subtypes do not appear to be independent predictors of inferior EFS or OS.[46,123] For instance, in a study from the U.K. Medical Research Council, the ETP ALL subgroup of patients had nonsignificantly inferior 5-year EFS rates compared with non-ETP patients (76% vs. 84%).[123] Similarly, in the COG AALL0434 [NCT00408005] trial, neither ETP nor near-ETP status had a statistically significant impact on EFS on multivariable analysis.[46] Based on these results, most ALL treatment groups do not change patient treatment based on ETP status.
Sources: 21 · 39 · 82 · 21 · 39 · 42 · 43 · 82 · 117 · 38 · 45 · 118 · 119 · 120 · 46 · 120 · 122 · 45 · 46 · 46 · 123 · 123 · AALL0434 [NCT00408005] · 46
T-ALL. T-ALL is defined by expression of the T-cell–associated antigens (cytoplasmic CD3, with CD7 plus CD2 or CD5) on leukemic blasts. T-ALL is frequently associated with a constellation of clinical features, including the following:[21,39,82] While not true historically, with appropriately intensive therapy, children with T-ALL now have an outcome approaching that of children with B-ALL.[21,39,42,43,82,117] There are few commonly accepted prognostic factors for patients with T-ALL. Conflicting data exist regarding the prognostic significance of presenting leukocyte counts in T-ALL.[38-45,118] The presence or absence of a mediastinal mass at diagnosis has no prognostic significance. In patients with a mediastinal mass, the rate of regression of the mass lacks prognostic significance.[119] Early T-cell precursor (ETP) ALL. ETP ALL, a distinct subset of childhood T-ALL, was initially defined by identifying T-ALL cases with gene expression profiles highly related to expression profiles for normal early T-cell precursors.[120] The subset of T-ALL cases identified by these analyses represented 13% of all cases, and are characterized by a distinctive immunophenotype (CD1a and CD8 negativity, with weak expression of CD5 and coexpression of stem cell or myeloid markers). Another subgroup of T-ALL, called near-ETP ALL, has a similar immunophenotype as ETP ALL, except with strong CD5 expression. This subtype represents approximately 15% of cases.[46] Initial reports describing ETP ALL suggested that this subset of patients has a poorer prognosis than other patients with T-ALL.[120-122] In addition, studies have reported that patients with ETP and near-ETP ALL have a slower early response and higher frequency of induction failure.[45,46] However, despite higher rates of end-induction MRD and induction failure in these patients, the ETP and near-ETP subtypes do not appear to be independent predictors of inferior EFS or OS.[46,123] For instance, in a study from the U.K. Medical Research Council, the ETP ALL subgroup of patients had nonsignificantly inferior 5-year EFS rates compared with non-ETP patients (76% vs. 84%).[123] Similarly, in the COG AALL0434 [NCT00408005] trial, neither ETP nor near-ETP status had a statistically significant impact on EFS on multivariable analysis.[46] Based on these results, most ALL treatment groups do not change patient treatment based on ETP status.
Male sex.
Older age.
Leukocytosis.
Mediastinal mass.
Myeloid antigen expression. Up to one-third of childhood ALL patients have leukemia cells that express myeloid-associated surface antigens. Myeloid-associated antigen expression appears to be associated with specific ALL subgroups, notably those with KMT2A rearrangements, ETV6::RUNX1, and BCR::ABL1.[124-126] Patients with B-ALL who have gene rearrangements involving ZNF384 also commonly show myeloid antigen expression.[127,128] No independent adverse prognostic significance exists for myeloid-surface antigen expression.[124,125] For information about leukemia of ambiguous lineage, see the 2016 WHO Classification of Acute Leukemias of Ambiguous Lineage section.
Sources: 124 · 126 · 127 · 128 · 124 · 125 · 2016 WHO Classification of Acute Leukemias of Ambiguous Lineage
Cytogenetics/genomic alterations
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
For information about B-ALL and T-ALL cytogenetics/genomics and gene polymorphisms in drug metabolic pathways, see the Cytogenetics/Genomics of Childhood ALL section.
Response to initial treatment
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The rapidity with which leukemia cells are eliminated after initiation of treatment and the level of residual disease at the end of induction are associated with long-term outcome. Because treatment response is influenced by the drug sensitivity of leukemic cells and host pharmacodynamics and pharmacogenomics,[129] early response has strong prognostic significance. Various ways of evaluating the leukemia cell response to treatment have been used, including the following:
Sources: 129
MRD determination in bone marrow at the end of induction (EOI) and end of consolidation (EOC).
Sources: MRD determination in bone marrow at the end of induction (EOI) and end of consolidation (EOC)
Day 7 and day 14 bone marrow responses.
Peripheral blood response to steroid prophase.
Peripheral blood response to multiagent induction therapy.
Sources: Peripheral blood response to multiagent induction therapy
Peripheral blood MRD before end of induction (day 8, day 15).
Sources: Peripheral blood MRD before end of induction (day 8, day 15)
Persistent leukemia at the end of induction (induction failure).
Sources: Persistent leukemia at the end of induction (induction failure)
MRD determination in bone marrow at the end of induction (EOI) and end of consolidation (EOC)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Morphological assessment of residual leukemia in blood or bone marrow is often difficult and is relatively insensitive. Traditionally, a cutoff of 5% blasts in the bone marrow (detected by light microscopy) has been used to determine remission status. This corresponds to a level of 1 in 20 malignant cells. To detect lower levels of leukemic cells in either blood or marrow, specialized techniques are required. Such techniques include polymerase chain reaction (PCR) assays, which determine unique Ig/T-cell receptor gene rearrangements and fusion transcripts produced by chromosome translocations, or flow cytometric assays, which detect leukemia-specific immunophenotypes. With these techniques, detection of as few as 1 leukemia cell in 100,000 normal cells is possible, and MRD at the level of 1 in 10,000 cells (1 × 10-4 or 0.01%) can be detected routinely.[130] Newer techniques involving high-throughput sequencing (HTS) of Ig/T-cell receptor gene rearrangements can increase sensitivity of MRD detection to 1 in 1 million cells (1 × 10-6 or 0.0001%).[131]
Multiple studies have demonstrated that EOI MRD is an important, independent predictor of outcome in children and adolescents with B-lineage ALL.[132-134] MRD response discriminates outcome in subsets of patients defined by age, leukocyte count, and cytogenetic abnormalities.[135] In general, patients with higher levels of EOI MRD have a poorer prognosis than do those with lower or undetectable levels.[130,132-134] However, the absolute risk of relapse associated with specific MRD levels varies by genetic subtype. For example, at any given level of detectable EOI MRD, patients with favorable cytogenetics, such as ETV6::RUNX1 or high hyperdiploidy, have a lower absolute risk of subsequent relapse than do other patients, while patients with high-risk cytogenetics have a higher absolute risk of subsequent relapse than do other patients.[136] This observation may have important implications when MRD is used to develop risk classification plans.
EOI MRD is used by almost all groups as a factor in determining the intensity of postinduction treatment. Patients found to have higher EOI MRD levels (typically >0.1% to 0.01%) are allocated to more intensive therapies.[130,133,137]; [138][Level of evidence B4]
Sources: 130 · 133 · 137 · 138 · Level of evidence B4
A study of 619 children with ALL compared the prognostic utility of MRD by flow cytometry with the more sensitive HTS assay. Using an EOI MRD cut point level of 0.01%, HTS identified approximately 30% more cases as positive (i.e., >0.01%). Patients identified as positive by HTS but negative by flow cytometry had an intermediate prognosis, compared with patients categorized as either positive or negative by both methods. Patients meeting the criteria for standard-risk ALL with undetectable MRD by HTS had an especially good prognosis (5-year EFS rate, 98.1%).[131]
Sources: 131
MRD levels obtained 10 to 12 weeks after the start of treatment (EOC) have also been shown to be prognostically important. Patients with high levels of EOC MRD have a significantly inferior EFS compared with other patients.[134,135,139] In one study by the Dutch Children's Oncology Group, patients with low but detectable EOC MRD (<0.05%, assessed by Ig/TR PCR assay) fared as well as those with nondetectable MRD at this time point, if their subsequent MRD assessments were negative. However, these patients did poorly if their subsequent MRD assessments remained detectable by the PCR assay.[140]
B-ALL. For patients with B-ALL, evaluating MRD at two time points (EOI and EOC) can identify the following three prognostically distinct patient subsets:[135]
Sources: 135
B-ALL. For patients with B-ALL, evaluating MRD at two time points (EOI and EOC) can identify the following three prognostically distinct patient subsets:[135]
Low or undetectable EOI MRD: Best prognosis.
High EOI MRD but low or negative EOC MRD: Intermediate prognosis.
High EOC MRD (week 10–12 of therapy): Worst prognosis. The prognostic impact of EOC MRD is modulated by NCI risk criteria. NCI high-risk patients with high EOC MRD have DFS rates lower than NCI standard-risk patients who have similar MRD levels at this time point.[139]
Sources: 139
T-ALL. There are fewer studies documenting the prognostic significance of MRD in patients with T-ALL. The United Kingdom Acute Lymphoblastic Leukaemia (UKALL) group reported that T-ALL patients with nondetectable EOI MRD had excellent outcomes, while those with very high MRD levels (>5%) at EOI had a poor prognosis. However, for all other T-ALL patients, an association between EOI MRD level and relapse risk was not found.[136] The DFCI ALL consortium also reported that T-ALL patients with very low EOI MRD (<10-4) had a very favorable outcome.[45] In the COG AALL0434 trial, high EOI MRD (>0.1%) was an independent predictor of inferior EFS and OS for patients with T-ALL on multivariable analysis.[46] In this trial, high EOI MRD was associated with inferior EFS in patients with non-ETP and near-ETP T-ALL, but not in those meeting the immunophenotypic definition of ETP.[46]Another study also indicated that MRD at a later time point may be more prognostically significant in T-ALL.[141] In the Associazione Italiana di Ematologia e Oncologia Pediatrica (AIEOP) ALL-BFM-2000 (NCT00430118) trial, MRD status at day 78 (week 12) was the most important predictor for relapse in patients with T-ALL.[141] Patients with detectable MRD at EOI who had negative MRD by day 78 generally had a favorable prognosis, similar to that of patients who achieved MRD-negativity at the earlier EOI time point.[141] In the COG AALL0434 trial, EOC MRD was evaluated in patients with T-ALL who had very high EOI MRD (>1%). High EOC MRD was associated with a markedly inferior outcome.[46] The COG AALL1231 study confirmed the prognostic significance of EOC marrow MRD for patients with T-ALL.[142]
Sources: 136 · 45 · 46 · 46 · 141 · ALL-BFM-2000 (NCT00430118) · 141 · 141 · 46 · 142
MRD measurements, in conjunction with other presenting features, have also been used to identify subsets of patients with an extremely low risk of relapse. The COG reported a very favorable prognosis (5-year EFS rate, 97% ± 1%) for patients with precursor B-cell phenotype, NCI standard risk age/leukocyte count, CNS1 status, and favorable cytogenetic abnormalities (either high hyperdiploidy with favorable trisomies or the ETV6::RUNX1 fusion) who had less than 0.01% MRD levels at both day 8 (from peripheral blood) and EOI (from bone marrow).[133] The excellent outcomes in patients with low MRD at the EOI were sustained for more than 10 years from diagnosis.[143]
Modifying therapy on the basis of MRD determination has been shown to improve outcome.
The UKALL2003 (NCT00222612) study demonstrated that reduction of therapy (i.e., one rather than two courses of delayed intensification) did not adversely impact outcome in non-high–risk patients with favorable EOI MRD.[22][Level of evidence B1] In a randomized controlled trial, the UKALL2003 study also demonstrated improved EFS for standard-risk and intermediate-risk patients who received augmented therapy if EOI MRD was greater than 0.01% (5-year EFS rates, 89.6% for augmented therapy vs. 82.8% for standard therapy).[144]
Sources: UKALL2003 (NCT00222612) · 22 · Level of evidence B1 · 144
The Dutch ALL10 trial stratified patients into the following three risk groups on the basis of MRD after the first month of treatment and after the second cycle of chemotherapy:[145][Level of evidence B4] Compared with previous trials conducted by the same group, therapy was less intensive for standard-risk patients but more intensive for moderate-risk and high-risk patients. The overall 5-year EFS rate (87%) and OS rate (92%) were superior to the previous Dutch studies.
Sources: 145 · Level of evidence B4
The Dutch ALL10 trial stratified patients into the following three risk groups on the basis of MRD after the first month of treatment and after the second cycle of chemotherapy:[145][Level of evidence B4] Compared with previous trials conducted by the same group, therapy was less intensive for standard-risk patients but more intensive for moderate-risk and high-risk patients. The overall 5-year EFS rate (87%) and OS rate (92%) were superior to the previous Dutch studies.
Standard risk (low MRD after the first month of treatment).
Moderate risk (high MRD after the first month of treatment, low MRD after the second cycle of chemotherapy).
High risk (high MRD after the second cycle of chemotherapy).
In the DFCI ALL Consortium 05-001 trial, B-ALL patients with high EOI MRD (defined as ≥1 × 10-3) were classified as very high risk regardless of other presenting characteristics. These patients received an intensified cytotoxic chemotherapy backbone. The 5-year DFS rate for these patients was 77%, significantly better than outcome for such patients on previous trials, when EOI MRD was not used to stratify therapy.[8]
Sources: 8
Day 7 and day 14 bone marrow responses
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Patients who have a rapid reduction in leukemia cells to less than 5% in their bone marrow within 7 or 14 days after the initiation of multiagent chemotherapy have a more favorable prognosis than patients who have slower clearance of leukemia cells from the bone marrow.[146] MRD assessments at the EOI therapy have generally replaced day 7 and day 14 morphological assessments as response to therapy prognostic indicators because the latter lose their prognostic significance in multivariate analysis once MRD is included in the analyses.[133,147]
Peripheral blood response to steroid prophase
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Patients with a reduction in peripheral blast count to less than 1,000/µL after a 7-day induction prophase with prednisone and one dose of intrathecal methotrexate (a good prednisone response) have a more favorable prognosis than patients whose peripheral blast counts remain above 1,000/µL (a poor prednisone response).[21] Poor prednisone response is observed in fewer than 10% of patients.[21,148] Treatment stratification for protocols of the Berlin-Frankfurt-Münster (BFM) clinical trials group historically were partially based on early response to the 7-day prednisone prophase (administered immediately before the initiation of multiagent remission induction). The current trial being conducted by that group still uses prednisone response to risk-stratify patients with T-ALL but not B-ALL.
Peripheral blood response to multiagent induction therapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Patients with persistent circulating leukemia cells at 7 to 10 days after the initiation of multiagent chemotherapy are at increased risk of relapse, compared with patients who have clearance of peripheral blasts within 1 week of therapy initiation.[149] Rate of clearance of peripheral blasts has been found to be of prognostic significance in both T-cell and B-lineage ALL.[149]
Peripheral blood MRD before EOI (day 8, day 15)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
MRD measured in peripheral blood obtained 1 week after the initiation of multiagent induction chemotherapy has also been evaluated as an early response-to-therapy prognostic factor.
In a COG study involving nearly 2,000 children with ALL, the presence of MRD in the peripheral blood at day 8 was associated with adverse prognosis. Increasing MRD levels were associated with a progressively poorer outcome.[133]
Sources: 133
In multivariate analysis, EOI MRD was the most powerful prognostic factor, but day 8 peripheral blood MRD maintained its prognostic significance, as did NCI risk group and the presence of favorable trisomies. A smaller study assessed the prognostic significance of peripheral blood MRD at day 15 after 1 week of a steroid prophase and 1 week of multiagent induction therapy.[150] This study also observed multivariate significance for peripheral blood MRD levels after 1 week of multiagent induction therapy.
Sources: 150
Both studies identified a group of patients who achieved low MRD levels after 1 week of multiagent induction therapy who had a low rate of subsequent treatment failure.
Persistent leukemia at the EOI (induction failure)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Nearly all children with ALL achieve complete morphological remission by the end of the first month of treatment. The presence of greater than 5% lymphoblasts by morphological assessment at the EOI phase is observed in 1% to 2% of children with ALL.[22,23,151-153]
Features associated with a higher risk of induction failure include the following:[153-155]
T-cell phenotype, especially the ETP phenotype.[46]
Sources: 46
Higher WBC at diagnosis for patients with B-ALL.
Older age.
Unfavorable biology.
Unfavorable biology.
KMT2A rearrangement.
BCR::ABL1 rearrangement (before the use of tyrosine kinase inhibitors).
Rearrangement of PDGFRB (most commonly EBF1::PDGFRB), commonly associated with the BCR::ABL1-like subtype.[153,156] These patients represent less than 1% of B-ALL cases in children but account for as much as 10% of induction failure cases.[153] In one retrospective study, 43 of 49 patients (88%) with PDGFRB fusions had EOI MRD levels greater than 1%.[157]
In a large retrospective study, the OS rate of patients with induction failure was only 32%.[151] However, there was significant clinical and biological heterogeneity. A relatively favorable outcome was observed in patients with B-ALL between the ages of 1 and 5 years without adverse cytogenetics (KMT2A rearrangement or BCR::ABL1). This group had a 10-year survival rate exceeding 50%, and hematopoietic stem cell transplant (HSCT) in first remission was not associated with a survival advantage compared with chemotherapy alone for this subset. Patients with the poorest outcomes (10-year survival rate, <20%) included those who were aged 14 to 18 years, or who had the BCR::ABL1 fusion or KMT2A rearrangement. B-ALL patients younger than 6 years and T-ALL patients (regardless of age) appeared to have better outcomes if treated with allogeneic HSCT after achieving CR than those who received further treatment with chemotherapy alone.[151] However, in the COG AALL0434 (NCT00408005) study, an advantage for HSCT in first CR for T-ALL patients with induction failure (defined as M3 marrow at EOI) was not observed. In this study, T-ALL patients were assigned to receive nelarabine during several postinduction treatment phases and high-dose methotrexate during the first interim maintenance phase. The 5-year EFS rate of these patients was 53.1%, with no significant difference between those who proceeded to HSCT in first CR (n = 20) and those who did not (n = 23) (P = .42).[158]
Sources: 151 · 151 · AALL0434 (NCT00408005) · 158
Flow cytometry versus morphology
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
MRD is now being integrated with morphological assessment into the response to induction therapy, on the basis of studies that showed that patients with MRD levels above 5%, despite morphological CR, had outcomes similar to patients with morphological induction failure.
In the UKALL2003 (NCT00222612) study, 59 of 3,113 patients (1.9%) had morphological induction failure.[153]
Sources: UKALL2003 (NCT00222612) · 153
In the UKALL2003 (NCT00222612) study, 59 of 3,113 patients (1.9%) had morphological induction failure.[153]
The 5-year EFS rate was 51%, and the OS rate was 58%.
2.3% of patients had a morphological remission but had MRD of ≥5% measured by real-time quantitative IgH–T-cell receptor PCR. This group had a 5-year EFS rate of 47%, similar to those with morphological induction failure.
The authors suggested that using both morphological and MRD criteria to define induction failure would more precisely identify patients with poor outcomes.
A study of 9,350 patients enrolled on COG clinical trials between 2004 and 2014 compared characteristics of patients and their outcomes categorized by morphology (M1 vs. M2/M3) and MRD status assessed by flow cytometry (<5% vs. ≥5%). Morphological remission (M1 status) was achieved for 98.6% of B-ALL patients and 93.8% of T-ALL patients at the EOI.[159]
Sources: 159
A study of 9,350 patients enrolled on COG clinical trials between 2004 and 2014 compared characteristics of patients and their outcomes categorized by morphology (M1 vs. M2/M3) and MRD status assessed by flow cytometry (<5% vs. ≥5%). Morphological remission (M1 status) was achieved for 98.6% of B-ALL patients and 93.8% of T-ALL patients at the EOI.[159]
Morphology and MRD were concordant in 97.4% of children. However, only 87.3% of T-ALL patients were M1 with MRD of <5%, while 97.8% of B-ALL patients were in concordant remission.
Approximately 20% of patients (40 of 202) with M2/M3 morphology had MRD of <5%. B-ALL patients with M2/M3 morphology but MRD of <5% had a 5-year OS rate of 72.7%, which was inferior to that of patients concordantly in remission (5-year OS rate, 93.8%) but superior to that of patients with M3 marrow (5-year OS rate, 43.4%).
Among B-ALL and T-ALL patients with M1 marrow, 0.9% of B-ALL patients and 6.9% of T-ALL patients had MRD of ≥5%. Their outcome was compared with that of patients with M1 marrow and MRD of <5% and are shown in Table 5 below.
Sources: Table 5
Table 5 shows that for children with B-ALL with M1 marrow and MRD of ≥5%, the 5-year EFS rate was significantly inferior to that of children concordantly in remission (59.1% vs. 87.1%) but was superior to that of children concordantly not in remission (M2 with MRD ≥5%: 5-year EFS rate, 39.1%).
Sources: Table 5
The impact on EFS for MRD of ≥5% for children with B-ALL in morphological remission was driven by NCI high-risk patients, as there was no significant difference in EFS between NCI standard-risk patients in morphological remission with or without MRD of ≥5%.
Inferior EFS rates were also observed for patients with T-ALL with M1 marrow and MRD of ≥5% compared with those in concordant remission (87.6% vs. 80.3%). However, outcome for T-ALL patients not in remission (whether by morphology or MRD or both) was superior to that of comparable patients with B-ALL.
Factors predictive of discordant MRD (≥5%) for patients in morphological remission at EOI included age 10 years and older, WBC count at presentation of 50,000/µL or higher, neutral or unfavorable cytogenetics, and ETP ALL (for patients with T-ALL).
| Outcome | M1/MRD <5% | P valueb | M1/MRD ≥5% | P valuec | M2/MRD ≥5% | |
|---|---|---|---|---|---|---|
| HR = high risk; MRD = minimal residual disease; SR = standard risk. | ||||||
| aAdapted from Gupta et al.[159] | ||||||
| bP value is comparing M1/MRD <5% with M1/MRD ≥5%. | ||||||
| cP value is comparing M1/MRD ≥5% with M2/MRD ≥5%. | ||||||
| Event-free survival rates: | ||||||
| B-ALL, overall | 87.1% ± 0.4% (n = 7,682) | <.0001 | 59.1% ± 6.5% (n = 66) | .009 | 39.1% ± 7.9% (n = 40) | |
| B-ALL, SR | 90.8% ± 0.4% (n = 5,000) | .25 | 85.9% ± 7.6% (n = 22) | .45 | 76.2% ± 15.2% (n = 9) | |
| B-ALL, HR | 80% ± 0.9% (n = 2,682) | <.0001 | 44.9% ± 8.3% (n = 44) | .05 | 29% ± 8.2% (n = 31) | |
| T-ALL | 87.6% ± 1.5% (n = 1,303) | .01 | 80.3% ± 7.3% (n = 97) | .13 | 62.7% ± 13.5% (n = 40) | |
| Overall survival rates: | ||||||
| B-ALL, overall | 93.8% ± 0.3% (n = 7,682) | <.0001 | 77.2% ± 5.6% (n = 66) | .01 | 59% ± 8.9% (n = 40) | |
| B-ALL, SR | 96.6% ± 0.3% (n = 5,000) | .24 | 95.5% ± 4.6% (n = 22 ) | .75 | 88.9% ± 12.1% (n = 9) | |
| B-ALL, HR | 88.4% ± 0.7% (n = 2,682) | <.0001 | 66.9% ± 8.3% (n = 44) | .06 | 51.4% ± 10.4% (n = 31) | |
| T-ALL | 91.9% ± 1.3% (n = 1,303) | .005 | 83.4% ± 6.8% (n = 97) | .34 | 76.7% ± 12.3% (n = 40) | |
Sources: 159
Prognostic (Risk) Groups
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
For decades, clinical trial groups studying childhood ALL have used risk classification schemes to assign patients to therapeutic regimens on the basis of their estimated risk of treatment failure. Initial risk classification systems used clinical factors such as age and presenting WBC count. Response-to-therapy measures were subsequently added, with some groups using early morphological bone marrow response (e.g., at day 8 or day 15) and with other groups using response of circulating leukemia cells to single-agent prednisone. Contemporary risk classification systems continue to use clinical factors such as age and presenting WBC count and incorporate cytogenetics and genomic lesions of leukemia cells at diagnosis (e.g., favorable and unfavorable translocations) and response to therapy based on detection of MRD at EOI (and in some cases at later time points).[141] The risk classification systems of the COG and the BFM groups are briefly described below.
Sources: 141
Children’s Oncology Group (COG) risk groups
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
In COG protocols, children with ALL are initially stratified into treatment groups (with varying degrees of risk of treatment failure) on the basis of a subset of prognostic factors, including the following:
Age.
WBC count at diagnosis.
Immunophenotype.
Cytogenetics/genomic alterations.
Presence of extramedullary disease.
Down syndrome.
Steroid pretreatment.
EFS rates exceed 85% in children meeting good-risk criteria (aged 1 to <10 years, WBC count <50,000/μL, and precursor B-cell immunophenotype). In children meeting high-risk criteria, EFS rates are approximately 75%.[4,57,148,160,161] Additional factors, including cytogenetic abnormalities and measures of early response to therapy (e.g., MRD levels in peripheral blood on day 8 and in bone marrow samples at the EOI), considered in conjunction with presenting age, WBC count, immunophenotype, the presence of extramedullary disease, and steroid pretreatment can identify patient groups for postinduction therapy with expected EFS rates ranging from less than 40% to more than 95%.[4,133]
Patients who are at very high risk of treatment failure include the following:[162-165]
Infants with KMT2A rearrangements.
Patients with hypodiploidy (<44 chromosomes).
Patients with initial induction failure.
Berlin-Frankfurt-Münster (BFM) risk groups
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Since 2000, risk stratification on BFM protocols has been based on treatment response criteria, as well as biology. Treatment response is assessed primarily via MRD measurements at two time points, EOI (time point 1, week 5) and end of the IB phase (similar to COG consolidation phase) at week 12 (time point 2). High MRD at both time points is defined as higher than 5 × 10-4.
The BFM defines 3 risk groups based on early response:[135]
Sources: 135
Standard risk: Patients who have negative MRD at both time points.
Intermediate risk: Patients who have high MRD at time point 1 and negative MRD at time point 2.
High risk: Patients with high MRD at time point 2. Patients with T-ALL with a poor response to the prednisone prophase are also considered high risk, regardless of subsequent MRD.
Biological factors used to stratify patients as high risk (regardless of MRD at either time point) include KMT2A::AFF1, TCF3::HLF, and hypodiploidy (<45 chromosomes). Patients with IKZF1-plus status (IKZF1 deletions that co-occurred with deletions in CDKN2A, CDKN2B, PAX5, or PAR1 in the absence of ERG deletion) [166] are considered high risk if they have high MRD at EOI, regardless of EOC MRD. Age, presenting leukocyte count, and CNS status at diagnosis do not factor into the current risk classification schema.
Sources: 166
Prognostic (risk) groups under clinical evaluation
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity.
Sources: AALL1731 (NCT03914625) · AALL1732 (NCT03959085) · Table 6 · Table 7 · Table 8 · Table 9 · Table 10 · NCT03792633
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity.
Age and presenting leukocyte count (using NCI risk-group criteria).[3]
Sources: 3
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity. › Age and presenting leukocyte count (using NCI risk-group criteria).[3]
NCI standard (low) risk: Includes children aged 1 year to <10 years with WBC <50,000/µL at the time of diagnosis.
NCI high risk: Includes children aged ≥10 years and/or children who have WBC ≥50,000/µL at the time of diagnosis.
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity.
Extramedullary disease (presence or absence of CNS and/or testicular leukemia).
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity. › Extramedullary disease (presence or absence of CNS and/or testicular leukemia).
CNS1: Absence of blasts on CSF cytospin preparation, regardless of the number of WBCs.
CNS2: Presence of <5 WBC/μL in CSF and cytospin positive for blasts; or traumatic LP, ≥5 WBC/μL, cytospin positive for blasts but negative by Steinherz/Bleyer algorithm.
CNS3 is divided and defined as follows:
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity. › Extramedullary disease (presence or absence of CNS and/or testicular leukemia). › CNS3 is divided and defined as follows:
CNS3a: <10 RBC/μL; ≥5 WBC/μL and cytospin positive for blasts.
CNS3b: ≥10 RBC/μL; ≥5 WBC/μL and positive by Steinherz/Bleyer algorithm.
CNS3c: Clinical signs of CNS leukemia (such as facial nerve palsy, brain/eye involvement or hypothalamic syndrome).
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity.
Genomic alterations in leukemia cells.
Day 8 peripheral blood MRD.
Day 29 bone marrow morphological response and MRD.
EOC MRD.
Steroid pretreatment.
Favorable cytogenetic features include the following:
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity. › Favorable cytogenetic features include the following:
Hyperdiploidy with double trisomies of chromosomes 4 and 10 (double trisomy); or
ETV6::RUNX1 fusion.
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity.
Unfavorable cytogenetic features include the following:
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity. › Unfavorable cytogenetic features include the following:
Hypodiploidy (<44 chromosomes or DNA index <0.81).
KMT2A rearrangements.
t(17;19)(q21-q22;p13.3) or resultant TCF3::HLF fusion transcript.
Intrachromosomal amplification of chromosome 21 (iAMP21); and
BCR::ABL1 (Ph+ or t(9;22)(q34;q11)). Patients with BCR::ABL1 ALL are treated on a separate clinical trial.
COG AALL1731 (NCT03914625) standard-risk and AALL1732 (NCT03959085) high-risk clinical trials: The COG classifies patients into six risk groups for patients with B-ALL (standard-risk favorable, standard-risk average, standard-risk high, high-risk favorable, high risk, and very high risk) on the basis of the following:Morphological assessment of early response in the bone marrow is no longer performed on days 8 and 15 of induction as part of risk stratification. Patients with T-cell phenotype are treated on separate trials and are not risk classified in this way.For patients with B-ALL, the definitions of favorable, unfavorable, and neutral cytogenetics are as follows:NCI standard-risk patients are divided into a highly favorable group (standard-risk favorable; 5-year DFS rate, >95%), a group with favorable outcome (standard-risk average; 5-year DFS rate, 90%–95%), and a group with a 5-year DFS rate below 90% (standard-risk high). Patients classified as standard-risk high receive postinduction backbone chemotherapy as per high-risk B-ALL regimens with intensified consolidation, interim maintenance, and reinduction therapy. Criteria for these three groups are provided in Table 6, Table 7, and Table 8 below.High-risk favorable B-ALL is defined by the characteristics in Table 9. These patients have an EFS rate higher than 90% on past COG clinical trials for high-risk patients. High-risk B-ALL is defined by the characteristics in Table 10. NCI standard-risk patients are elevated to high-risk status based on steroid pretreatment and CNS and/or testicular involvement.NCI high-risk patients with EOC marrow MRD ≥0.01% are classified as very high risk and are eligible for a chimeric antigen receptor (CAR) T-cell clinical trial in first remission (NCT03792633).Patients with B-ALL and Down syndrome are classified into risk groups similar to other children, but Down syndrome patients classified as high risk receive a treatment regimen that is modified to reduce toxicity.
Neutral cytogenetics: Lacking favorable and unfavorable cytogenetic features.
| NCI Risk Group | CNS Stage | Steroid Pretreatmenta | Favorable Genetics (ETV6::RUNX1 or DT) | PB MRD Day 8 | BM MRD Day 29 |
|---|---|---|---|---|---|
| BM = bone marrow; CNS = central nervous system; DT = double trisomy; MRD = minimal residual disease; NCI = National Cancer Institute; PB = peripheral blood; SR = standard risk. | |||||
| aWithin one month prior to diagnosis. | |||||
| SR | 1, 2 | None | Yes | <1% | <0.01% |
| NCI Risk Group | CNS Stage | ETV6::RUNX1 | DT | Neutral Cytogenetics | PB MRD Day 8 | BM MRD Day 29 |
|---|---|---|---|---|---|---|
| BM = bone marrow; CNS = central nervous system; DT = double trisomy; MRD = minimal residual disease; NCI = National Cancer Institute; PB = peripheral blood; SR = standard risk. | ||||||
| SR | 1, 2 | Yes to either | No | ≥1% | <0.01% | |
| SR | 1, 2 | No | Yes | No | Any | ≥0.01 to <0.1% |
| SR | 1 | No | No | Yes | Any | <0.01% |
| NCI Risk Group | CNS Stage | ETV6::RUNX1 | DT | Neutral Cytogenetics | Unfavorable Cytogenetics | PB MRD Day 8 | BM MRD Day 29 |
|---|---|---|---|---|---|---|---|
| BM = bone marrow; CNS = central nervous system; DT = double trisomy; MRD = minimal residual disease; NCI = National Cancer Institute; PB = peripheral blood; SR = standard risk. | |||||||
| SR | 1, 2 | Yes | No | No | No | Any | ≥0.01% |
| SR | 1, 2 | No | Yes | No | No | Any | ≥0.1% |
| SR | 1 | No | No | Yes | No | Any | ≥0.01% |
| SR | 2 | No | No | Yes | No | Any | Any |
| SR | 1, 2 | No | No | No | Yes | Any | Any |
| NCI Risk Group | Age (y) | CNS Status | Testicular Leukemia | Steroid Pretreatment | Favorable Genetics (ETV6::RUNX1 or DT) | Bone marrow MRD EOI |
|---|---|---|---|---|---|---|
| HR | <10 | 1 | None | ≤24 hoursa | Yes | <0.01% |
| CNS = central nervous system; DT = double trisomy; EOI = end of induction; HR = high risk; MRD = minimal residual disease; NCI = National Cancer Institute. | ||||||
| aWithin two weeks of diagnosis. | ||||||
| NCI Risk Group | Age (y) | CNS and/or Testicular Leukemia | Steroid Pretreatment | Cytogenetics | Bone marrow MRD EOI | Bone marrow MRD EOC |
|---|---|---|---|---|---|---|
| CNS = central nervous system; EOC = end of consolidation; EOI = end of induction; HR = high risk; MRD = minimal residual disease; N/A = not applicable; NCI = National Cancer Institute; SR = standard risk. | ||||||
| aCNS3. | ||||||
| bPhiladelphia chromosome–positive (Ph+) ALL is excluded. | ||||||
| cOnly subjects with EOI bone marrow MRD ≥0.01% will have a bone marrow MRD assessment at EOC. | ||||||
| dWithin 2 weeks of diagnosis. | ||||||
| eCNS2 or CNS3. | ||||||
| SR | <10 | Yesa | Any | Anyb | Any | <1%c |
| SR | <10 | No | >24 hoursd | Anyb | Any | <1%c |
| HR | ≥10 | Any | Any | Anyb | <0.01% | N/A |
| HR | <10 | Yese | Any | Anyb | <0.01% | N/A |
| HR | <10 | No | >24 hoursd | Anyb | <0.01% | N/A |
| HR | <10 | No | ≤24 hoursd | Neutral/unfavorableb | <0.01% | N/A |
| HR | Any | Any | Any | Anyb | ≥0.01% | <0.01% |
Treatment Option Overview for Childhood ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Phases of Therapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Treatment for children with acute lymphoblastic leukemia (ALL) is typically divided into the following phases:
Remission induction chemotherapy (at the time of diagnosis).
Sources: Remission induction chemotherapy
Postinduction therapy (after achieving complete remission).
Sources: Postinduction therapy
Postinduction therapy (after achieving complete remission).
Consolidation/intensification therapy.
Maintenance therapy.
Sources: Maintenance therapy
Sanctuary Sites
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Historically, certain extramedullary sites have been considered sanctuary sites (i.e., anatomic spaces that are poorly penetrated by many of the orally and intravenously administered chemotherapy agents typically used to treat ALL). The two most important sanctuary sites in childhood ALL are the central nervous system (CNS) and the testes. Successful treatment of ALL requires therapy that effectively addresses clinical or subclinical involvement of leukemia in these extramedullary sanctuary sites.
Central nervous system (CNS)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
At diagnosis, approximately 3% of patients have CNS3 disease (defined as cerebrospinal fluid specimen with ≥5 white blood cells/μL with lymphoblasts and/or the presence of cranial nerve palsies). However, unless specific therapy is directed toward the CNS, most children will eventually develop overt CNS leukemia whether or not lymphoblasts were detected in the spinal fluid at initial diagnosis. CNS-directed treatments include intrathecal chemotherapy, CNS-directed systemic chemotherapy, and cranial radiation. Some or all of these treatments are included in current regimens for ALL. For more information, see the CNS-Directed Therapy for Childhood ALL section.
Testes
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Overt testicular involvement at the time of diagnosis occurs in approximately 2% of males. In early ALL trials, testicular involvement at diagnosis was an adverse prognostic factor. With more aggressive initial therapy, however, the prognostic significance of initial testicular involvement is unclear.[1,2] The role of radiation therapy for testicular involvement is also unclear. A study from St. Jude Children's Research Hospital suggests that a good outcome can be achieved with aggressive conventional chemotherapy without radiation.[1] The Children's Oncology Group has also adopted this strategy for boys with testicular involvement that resolves completely during induction chemotherapy.
Special Considerations for the Treatment of Children With ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The treatment of children and adolescents with acute lymphoblastic leukemia (ALL) entails complicated risk assignment, extensive therapies, and intensive supportive care (e.g., transfusions; management of infectious complications; and emotional, financial, and developmental support). Because of these factors, the evaluation and treatment of these patients are best coordinated by a multidisciplinary team in cancer centers or hospitals with all of the necessary pediatric supportive care facilities.[1] This multidisciplinary team approach incorporates the skills of the following pediatric specialists and others to ensure that children receive treatment, supportive care, and rehabilitation to achieve optimal survival and quality of life:
Sources: 1
Primary care physicians.
Pediatric medical oncologists and hematologists.
Pediatric surgeons.
Pathologists.
Pediatric radiation oncologists.
Pediatric intensivists.
Rehabilitation specialists.
Pediatric oncology nurses.
Social workers.
Child-life professionals.
Psychologists.
Nutritionists.
For specific information about supportive care for children and adolescents with cancer, see the summaries on Supportive and Palliative Care.
Sources: Supportive and Palliative Care
The American Academy of Pediatrics has outlined guidelines for pediatric cancer centers and their role in the treatment of children and adolescents with cancer.[1] Treatment of childhood ALL typically involves chemotherapy given for 2 to 3 years. Because myelosuppression and generalized immunosuppression are anticipated consequences of leukemia and chemotherapy treatment, adequate facilities must be immediately available for both hematological support and treatment of infections and other complications throughout all phases of therapy. Approximately 1% to 3% of patients die during the remission induction phase, and another 1% to 3% die after having achieved complete remission from treatment-related complications.[2-6] It is important that the clinical centers and the specialists directing the patient’s care maintain contact with the referring physician in the community. Strong lines of communication optimize any urgent or interim care required when the child is at home.
Clinical trials are generally available for children with ALL, with specific protocols designed for children at standard (low) risk of treatment failure and for children at higher risk of treatment failure. Clinical trials for children with ALL are generally designed to compare standard therapy for a particular risk group with a potentially better treatment approach that may improve survival and/or diminish toxicities associated with the standard treatment regimen. Other types of clinical trials test novel therapies when there is no standard therapy for a cancer diagnosis. Many of the therapeutic innovations that produced increased survival rates in children with ALL were achieved through clinical trials, and it is appropriate for children and adolescents with ALL to be offered participation in a clinical trial. Information about ongoing clinical trials is available from the NCI website.
Sources: NCI website
Risk-based treatment assignment is an important therapeutic strategy for children with ALL. This approach allows children who historically have a very good outcome to be treated with less intensive therapy and to be spared more toxic treatments, while children with a historically lower probability of long-term survival receive more intensive therapy that may increase their chance of cure. For more information about clinical and laboratory features that have shown prognostic value, see the Risk-Based Treatment Assignment section.
Sources: Risk-Based Treatment Assignment
Treatment of Newly Diagnosed Childhood ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Standard Induction Treatment Options for Newly Diagnosed ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Standard treatment options for newly diagnosed childhood acute lymphoblastic leukemia (ALL) include the following:
Chemotherapy.
Remission induction chemotherapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The goal of the first phase of therapy (remission induction) is to induce a complete remission (CR). This induction phase typically lasts 4 weeks. Overall, approximately 98% of patients with newly diagnosed B-ALL achieve CR by the end of this phase, with somewhat lower rates in infants and in noninfant patients with T-ALL or high presenting leukocyte counts.[1-5]
Induction chemotherapy typically consists of the following drugs, with or without an anthracycline (either doxorubicin or daunorubicin):
Vincristine.
Corticosteroid (either prednisone or dexamethasone).
Asparaginase.
Intrathecal chemotherapy.
The Children's Oncology Group (COG) protocols administer a three-drug induction (vincristine, corticosteroid, and pegaspargase) to National Cancer Institute (NCI) standard-risk B-ALL patients and a four-drug induction (vincristine, corticosteroid, and pegaspargase plus anthracycline) to NCI high-risk B-ALL and all T-ALL patients. Other groups use a four-drug induction for all patients.[1-3]
Corticosteroid therapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Many current regimens use dexamethasone instead of prednisone during remission induction and later phases of therapy, although controversy exists as to whether dexamethasone benefits all subsets of patients. Some trials also suggest that dexamethasone during induction may be associated with more toxicity than prednisone, including higher rates of infection, myopathy, and behavioral changes.[1,6-8] The COG reported that dexamethasone during induction was associated with a higher risk of osteonecrosis in older children (aged >10 years),[8] although this finding has not been confirmed in other randomized studies.[1,7]
Evidence (dexamethasone vs. prednisone during induction):
The Children's Cancer Group conducted a randomized trial that compared dexamethasone and prednisone in standard-risk B-ALL patients receiving a three-drug induction without an anthracycline.[6]
Sources: 6
The Children's Cancer Group conducted a randomized trial that compared dexamethasone and prednisone in standard-risk B-ALL patients receiving a three-drug induction without an anthracycline.[6]
Dexamethasone was associated with a superior event-free survival (EFS).
Dexamethasone was associated with a higher frequency of reversible steroid myopathy and hyperglycemia. No significant differences in rates of infection during induction were observed between the two randomized arms.
Another randomized trial that included both standard-risk and high-risk patients was conducted by the United Kingdom Medical Research Council (MRC).[7]
Sources: 7
Another randomized trial that included both standard-risk and high-risk patients was conducted by the United Kingdom Medical Research Council (MRC).[7]
The trial demonstrated that dexamethasone was associated with a more favorable outcome than prednisolone in all patient subgroups.
Patients who received dexamethasone had a significantly lower incidence of both central nervous system (CNS) and non-CNS relapses than did patients who received prednisolone.
Dexamethasone was associated with a higher incidence of steroid-associated behavioral problems and myopathy, but an excess risk of osteonecrosis was not observed. There was no difference in induction death rates between the randomized groups.
The Associazione Italiana di Ematologia e Oncologia Pediatrica (AIEOP) ALL-BFM-2000 (NCT00430118) trial randomly assigned 3,720 patients to receive either dexamethasone (10 mg/m2/d) or prednisone (60 mg/m2/d) during multiagent remission induction (including an anthracycline for all patients) after a 7-day prednisone prophase.[9]
Sources: ALL-BFM-2000 (NCT00430118) · 9
The Associazione Italiana di Ematologia e Oncologia Pediatrica (AIEOP) ALL-BFM-2000 (NCT00430118) trial randomly assigned 3,720 patients to receive either dexamethasone (10 mg/m2/d) or prednisone (60 mg/m2/d) during multiagent remission induction (including an anthracycline for all patients) after a 7-day prednisone prophase.[9]
Dexamethasone was associated with higher incidence of life-threatening events (primarily infections), resulting in a significantly higher induction death rate (2.5% for dexamethasone vs. 0.9% for prednisone; P = .00013).
There was no difference in rates of osteonecrosis between the randomized groups.
The 5-year cumulative incidence of relapse was significantly lower with dexamethasone (11% vs. 16%; P < .0001), resulting in superior 5-year EFS rates (84% for dexamethasone vs. 81% for prednisone, P = .024) despite the increased induction death rate.
No difference in overall survival (OS) was observed based on steroid randomization, although the study was not sufficiently powered to detect small differences in OS.
In a predefined subgroup analysis, a survival benefit was observed with dexamethasone treatment in patients with T-ALL and a good response to the prednisone prophase (5-year OS rates, 91% with dexamethasone vs. 83% with prednisone, P = .036).
The COG conducted a randomized trial of dexamethasone and prednisone in NCI high-risk B-ALL patients.[8] Patients were randomly assigned to receive 14 days of dexamethasone or 28 days of prednisone during a four-drug induction (with an anthracycline). This trial also included a randomized comparison of high-dose and escalating-dose methotrexate during the interim maintenance phase.
Sources: 8
The COG conducted a randomized trial of dexamethasone and prednisone in NCI high-risk B-ALL patients.[8] Patients were randomly assigned to receive 14 days of dexamethasone or 28 days of prednisone during a four-drug induction (with an anthracycline). This trial also included a randomized comparison of high-dose and escalating-dose methotrexate during the interim maintenance phase.
Dexamethasone was associated with a higher rate of infection, but there was no difference in the induction death rate when comparing dexamethasone and prednisone.
For patients who were younger than 10 years at diagnosis, there was a significant interaction between the corticosteroid and methotrexate randomizations. However, the best outcome for this group of patients was observed in those who received both dexamethasone during induction and high-dose methotrexate during interim maintenance.
The corticosteroid randomization was closed early for patients aged 10 years or older at diagnosis because of excessive rates of osteonecrosis in patients randomly assigned to dexamethasone. However, there was no EFS benefit associated with dexamethasone in these older patients (5-year EFS rates of 73.1% with dexamethasone and 73.9% with prednisone; P = .78)
The ratio of dexamethasone to prednisone dose used may influence outcome. Studies in which the dexamethasone to prednisone ratio was 1:5 to 1:7 have shown a better result for dexamethasone, while studies that used a 1:10 ratio have shown similar outcomes.[10]
Sources: 10
Asparaginase
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Several forms of asparaginase have been used in the treatment of children with ALL, including the following:
Pegaspargase (PEG-asparaginase).
Sources: Pegaspargase (PEG-asparaginase)
Calaspargase pegol.
Sources: Calaspargase pegol
Asparaginase Erwinia chrysanthemi (Erwinia L-asparaginase).
Sources: Asparaginase Erwinia chrysanthemi (Erwinia L-asparaginase)
Native Escherichia coli (E. coli) L-asparaginase (unavailable in the United States, but still available in other countries).
Pegaspargase (PEG-asparaginase)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Pegaspargase is a form of L-asparaginase in which the E. coli–derived enzyme is modified by the covalent attachment of polyethylene glycol. It is commonly used during both induction and postinduction phases of treatment in newly diagnosed patients treated in Western Europe. Pegaspargase is not available in the United States, but it is still available in other countries.
Pegaspargase may be given either intramuscularly (IM) or intravenously (IV).[11] Pharmacokinetics and toxicity profiles are similar for IM and IV pegaspargase administration.[11] There is no evidence that IV administration of pegaspargase is more toxic than IM administration.[11-13]
Pegaspargase has a much longer serum half-life than native E. coli L-asparaginase, producing prolonged asparagine depletion after a single injection.[14]
Sources: 14
Serum asparaginase enzyme activity levels of more than 0.1 IU/mL have been associated with serum asparagine depletion. Studies have shown that a single dose of pegaspargase given either IM or IV as part of multiagent induction results in serum enzyme activity of more than 0.1 IU/mL in nearly all patients for at least 2 to 3 weeks.[11,12,15,16] In one study of 54 NCI high-risk patients conducted by the COG, plasma asparaginase activity as low as 0.02 IU/mL was associated with serum asparagine depletion. Using that cutoff value, it was estimated that 96% of patients maintained the therapeutic effect (plasma asparagine depletion) for 22 to 29 days after a single pegaspargase dose of 2,500 IU/m2.[17] In one randomized study, higher doses of pegaspargase (3,500 IU/m2) did not improve outcome when compared with standard doses (2,500 IU/m2).[18][Level of evidence A1]
In another study, doses of pegaspargase were reduced in an attempt to decrease toxicity.[19] While lower doses were successful in maintaining appropriate asparaginase levels of more than 0.1 IU/mL, the frequency of asparaginase-related toxicities was similar to the frequency of toxicities reported in previous studies that used higher doses of pegaspargase. This study did not report on the impact of lower doses of pegaspargase on EFS.
Sources: 19
Evidence (use of pegaspargase versus native E. coli L-asparaginase):
A randomized comparison of IV pegaspargase versus IM native E. coli asparaginase was conducted. Each agent was administered for a 30-week period after the achievement of CR.[13][Level of evidence A3]
Sources: 13 · Level of evidence A3
A randomized comparison of IV pegaspargase versus IM native E. coli asparaginase was conducted. Each agent was administered for a 30-week period after the achievement of CR.[13][Level of evidence A3]
Serum asparaginase activity (SAA) levels were significantly higher with IV pegaspargase and exceeded goal therapeutic levels (>0.1 IU/mL) in nearly all patients throughout the 30-week period.
There was no significant difference in EFS and OS between the randomized arms.
There was no difference in rates of asparaginase-related toxicities, including hypersensitivity, pancreatitis, and thromboembolic complications.
Similar outcome and similar rates of asparaginase-related toxicities were observed for both groups of patients.
IV pegaspargase was associated with less treatment-related anxiety, as assessed by patient and parent surveys.
Another randomized trial of patients with standard-risk ALL assigned patients to receive either pegaspargase or native E. coli asparaginase during induction and in each of two delayed intensification courses.[15]
Sources: 15
Another randomized trial of patients with standard-risk ALL assigned patients to receive either pegaspargase or native E. coli asparaginase during induction and in each of two delayed intensification courses.[15]
A single dose of pegaspargase given in conjunction with vincristine and prednisone during induction therapy appeared to have similar activity and toxicity as nine doses of IM E. coli L-asparaginase (3 times a week for 3 weeks).[15]
Sources: 15
The use of pegaspargase was associated with more rapid blast clearance and a lower incidence of neutralizing antibodies.
Patients with an allergic reaction to pegaspargase are typically switched to Erwinia L-asparaginase. A COG analysis investigated the deleterious effect on disease-free survival (DFS) of early discontinuation of treatment with pegaspargase in patients with high-risk B-ALL. The study found that the adverse effect on outcome could be reversed with the use of Erwinia L-asparaginase to complete the planned course of asparaginase therapy.[20][Level of evidence C2] Measurement of SAA levels after a mild or questionable reaction to pegaspargase may help to differentiate patients for whom the switch to Erwinia is indicated (because of inadequate SAA) versus those for whom a change in preparation may not be necessary.[21,22]
Sources: 20 · Level of evidence C2 · 21 · 22
Evidence (adverse prognostic impact of early discontinuation of pegaspargase or silent inactivation of asparaginase):
Several studies have identified a subset of patients who experience silent inactivation of asparaginase, which is defined as the absence of therapeutic SAA levels without overt allergy.[23,24] Determination of the optimal frequency of pharmacokinetic monitoring for pegaspargase-treated patients, and whether such screening impacts outcome, awaits further investigation.
Several studies have identified a subset of patients who experience silent inactivation of asparaginase, which is defined as the absence of therapeutic SAA levels without overt allergy.[23,24] Determination of the optimal frequency of pharmacokinetic monitoring for pegaspargase-treated patients, and whether such screening impacts outcome, awaits further investigation.
In a trial conducted by the Dana-Farber Cancer Institute (DFCI) Consortium, 12% of patients who were initially treated with native E.coli L-asparaginase demonstrated silent inactivation. These patients had a superior EFS if their asparaginase preparation was changed.[24]
Sources: 24
Patients who were treated with pegaspargase appear to have lower levels of silent inactivation (<10%).[13,23,25]
A report from the COG included 8,196 patients with newly diagnosed B-ALL who were enrolled between 2004 to 2011.[20][Level of evidence C2]
Sources: 20 · Level of evidence C2
A report from the COG included 8,196 patients with newly diagnosed B-ALL who were enrolled between 2004 to 2011.[20][Level of evidence C2]
The cumulative incidence of pegaspargase discontinuation (because of toxicity) was 12.2% in NCI standard-risk patients and 25.4% in NCI high-risk patients.
In multivariable analysis, NCI high-risk patients who discontinued pegaspargase early had inferior DFS (hazard ratio [HR], 1.5; P = .002) than did those who received all prescribed doses. For NCI standard-risk patients, there was no impact of pegaspargase discontinuation on DFS, except in patients with slow-early response who received intensified postinduction therapy (HR, 1.7; P = .03).
NCI high-risk patients who discontinued pegaspargase but then switched to Erwinia asparaginase and received all subsequent intended doses, did not have an increased risk of relapse (HR, 1.1; P = .69).
An analysis of 1,115 non–high-risk ALL patients from the Nordic Society of Pediatric Hematology and Oncology (NOPHO) ALL2008 protocol reported the following:[25]
Sources: 25
An analysis of 1,115 non–high-risk ALL patients from the Nordic Society of Pediatric Hematology and Oncology (NOPHO) ALL2008 protocol reported the following:[25]
255 patients received a truncated asparaginase course because of toxicity, and 46 patients had evidence of silent inactivation on therapeutic drug monitoring.
The 7-year cumulative incidence of relapse was 11.1% in the 301 patients who received a truncated asparaginase course, compared with 6.7% in the remaining 814 patients who received the planned courses (HR, 1.73; P = .03).
In a Cox model, suboptimal asparaginase treatment (because of either truncated pegaspargase or silent inactivation) was significantly associated with a higher relapse risk (HR, 1.69; P = 0.03).
In an attempt to decrease hypersensitivity reactions to pegaspargase, the Dutch Childhood Oncology Group-ALL11 protocol randomly assigned patients to receive either continuous or noncontinuous dosing after induction therapy. The occurrence of inactivating hypersensitivity reactions was seven times lower and antibody levels were significantly lower in the continuous-dosing arm. There was no difference in total number of asparaginase toxicities or the 5-year incidences of relapse, death, or disease-free survival between the treatment arms.[26]
Sources: 26
Calaspargase pegol
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Calaspargase pegol is another formulation of pegylated asparaginase that is also available for the treatment of children and adolescents with ALL.[27] This formulation is similar in structure to pegaspargase, except with a different linker between the L-asparaginase enzyme and the PEG moiety, resulting in a longer half-life.[28,29]
Evidence (calaspargase pegol vs. pegaspargase):
In a COG study, 165 patients with high-risk B-ALL were randomly assigned to receive either calaspargase pegol or pegaspargase during the induction phase of ALL therapy.[28]
Sources: 28
In a COG study, 165 patients with high-risk B-ALL were randomly assigned to receive either calaspargase pegol or pegaspargase during the induction phase of ALL therapy.[28]
The mean half-life of calaspargase pegol was approximately 2.5 times longer than pegaspargase.
The total systemic exposure to calaspargase pegol was greater than for pegaspargase.
Twenty-five days after a dose of calaspargase pegol, 95% of patients maintained an asparaginase level higher than 0.1 IU/mL, compared with 28% of patients who received pegaspargase.
Evidence of end-induction minimal residual disease (MRD) negativity was similar between the two drugs (74% and 72%).
The toxicity profile of the two drugs was similar.
In a DFCI trial of calaspargase pegol in patients with newly diagnosed ALL, all patients received one dose of either calaspargase pegol or pegaspargase as part of induction therapy. After induction, 230 patients were randomly assigned to receive either calaspargase pegol every 3 weeks (10 doses) or pegaspargase every 2 weeks (15 doses).[29]
Sources: 29
In a DFCI trial of calaspargase pegol in patients with newly diagnosed ALL, all patients received one dose of either calaspargase pegol or pegaspargase as part of induction therapy. After induction, 230 patients were randomly assigned to receive either calaspargase pegol every 3 weeks (10 doses) or pegaspargase every 2 weeks (15 doses).[29]
At day 25 after the induction dose, 88% of patients who received calaspargase pegol had an asparaginase level higher than 0.1 IU/mL, compared with 17% of patients who received pegaspargase.
There was no difference in end-of-induction MRD.
There was no difference in the frequency of toxicities (37%).
The 5-year EFS rates were similar for calaspargase pegol and pegaspargase (88.1% vs. 84.9%).
Calaspargase pegol has only been approved for use in the United States for patients younger than 22 years.
Asparaginase Erwinia chrysanthemi (Erwinia L-asparaginase)
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Erwinia L-asparaginase is typically used in patients who have experienced an allergy to native E. coli or pegaspargase.
The half-life of Erwinia L-asparaginase (0.65 days) is much shorter than that of native E. coli (1.2 days) or pegaspargase (5.7 days).[14] If Erwinia L-asparaginase is used, the shorter half-life of the Erwinia preparation requires more frequent administration to achieve adequate asparagine depletion.
Sources: 14
Evidence (increased dose frequency of Erwinia L-asparaginase needed to achieve goal therapeutic effect):
A COG trial demonstrated that IM Erwinia L-asparaginase given three times a week to patients with an allergy to pegaspargase leads to therapeutic serum asparaginase enzyme activity levels (defined as a level ≥0.1 IU/mL).[30]
Sources: 30
A COG trial demonstrated that IM Erwinia L-asparaginase given three times a week to patients with an allergy to pegaspargase leads to therapeutic serum asparaginase enzyme activity levels (defined as a level ≥0.1 IU/mL).[30]
On this trial, 96% of children achieved a level of 0.1 IU/mL or more at 2 days after a dose of Erwinia L-asparaginase and 85% did so at 3 days after a dose.
A trial of IV Erwinia L-asparaginase given on a Monday-Wednesday-Friday schedule to patients with an allergy to pegaspargase demonstrated therapeutic serum asparaginase enzyme activity (defined as ≥0.1 IU/mL) in 83% of patients 48 hours after a dose but in only 43% of patients 72 hours after a dose.[31]
Sources: 31
A trial of IV Erwinia L-asparaginase given on a Monday-Wednesday-Friday schedule to patients with an allergy to pegaspargase demonstrated therapeutic serum asparaginase enzyme activity (defined as ≥0.1 IU/mL) in 83% of patients 48 hours after a dose but in only 43% of patients 72 hours after a dose.[31]
If IV Erwinia is given on a Monday-Wednesday-Friday schedule, the authors suggest that 72-hour nadir enzyme activity levels be monitored to ensure therapeutic levels.
A recombinant form of Erwinia L-asparaginase, asparaginase erwinia chrysanthemi (recombinant)-rywn, was studied in a phase II/III COG trial. When it was given on a Monday (25 mg/m2), Wednesday (25 mg/m2), and Friday (50 mg/m2) schedule for six doses, the proportion of patients who achieved asparaginase levels of 0.1 IU/mL or greater was 90% at 72 hours (44 of 49 patients) and 96% at 48 hours (47 of 49 patients). The safety profile was comparable with other forms of asparaginase.[32] In 2022, the U.S. Food and Drug Administration approved asparaginase erwinia chrysanthemi (recombinant)-rywn for IM use in children and adults with ALL on the Monday, Wednesday, and Friday schedule used in the COG trial.
Sources: 32
Anthracycline use during induction
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The COG protocols administer a three-drug induction (vincristine, corticosteroid, and pegaspargase) to NCI standard-risk B-ALL patients and a four-drug induction (vincristine, corticosteroid, and pegaspargase plus an anthracycline) to NCI high-risk B-ALL and all T-ALL patients. Other groups use a four-drug induction for all patients.[1-3]
In induction regimens that include an anthracycline, either daunorubicin or doxorubicin are typically used. In a randomized trial comparing the two agents during induction, there were no differences in early response measures, including reduction in peripheral blood blast counts during the first week of therapy, day 15 marrow morphology, and end-induction MRD levels.[33][Level of evidence B3]
Sources: 33 · Level of evidence B3
Response to remission induction chemotherapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
More than 95% of children with newly diagnosed ALL will achieve a CR within the first 4 weeks of treatment. Of those who fail to achieve CR within the first 4 weeks, approximately one-half will experience a toxic death during the induction phase (usually caused by infection) and the other half will have resistant disease (persistent morphological leukemia).[34-36]; [37][Level of evidence C1]
Sources: 34 · 36 · 37 · Level of evidence C1
Remission is classically defined as an end-induction bone marrow examination by routine microscopic cytomorphology with fewer than 5% lymphoblasts at the end of induction (M1). The Ponte de Legno consortium includes approximately 15 large national and international cooperative groups devoted to the study and treatment of childhood ALL. This group published a consensus definition of complete remission, as follows:[38]
Sources: 38
Achievement of MRD levels of less than 1% and/or M1 cytomorphology.
Achievement of MRD levels of less than 1% and/or M1 cytomorphology.
MRD is the gold standard and takes precedence over cytomorphology.
MRD is determined by either flow cytometry or polymerase chain reaction techniques.
Resolution of extramedullary disease, assessed no earlier than the end of induction.
Most patients with persistence of morphologically detectable leukemia at the end of the 4-week induction phase have a poor prognosis and may benefit from an allogeneic hematopoietic stem cell transplant (HSCT) once CR is achieved.[4,39,40] In a retrospective study of 1,041 patients with persistent disease after induction therapy (induction failure) who were treated between 1985 and 2000, the 10-year OS rate was 32%.[41] A trend for superior outcome with allogeneic HSCT compared with chemotherapy alone was observed in patients with T-cell phenotype (any age) and B-ALL patients older than 6 years. B-ALL patients who were aged 1 to 5 years at diagnosis and did not have any adverse cytogenetic abnormalities (KMT2A rearrangement, BCR::ABL1) had a relatively favorable prognosis, without any advantage in outcome with the utilization of HSCT compared with chemotherapy alone.[41]
A follow-up retrospective study reported the outcomes of 325 children and adolescents with T-ALL and initial induction failure who were treated between 2000 and 2018.[42] The 10-year OS rate was 54.7%, which was significantly better than the rates of patients in historical cohorts who were treated between 1985 and 2000 (10-year OS rate, 27.6%). Complete remission was eventually achieved in 93% of patients with T-ALL and initial induction failure. Of the patients who achieved complete remission, 72% underwent HSCT. Adjusting for time to transplant, the 10-year OS rate was 66.2% for these patients, compared with 50.8% for those who did not undergo transplants.
Sources: 42
The incorporation of nelarabine may be of value for patients with T-ALL and have induction failure. The COG AALL0434 (NCT00408005) study included 43 patients with more than 25% blasts in an end-induction bone marrow aspirate. Of these patients, 23 patients were nonrandomly assigned to therapy that included high-dose methotrexate and nelarabine as part of a multidrug regimen, and 20 patients underwent allogeneic transplant. The 5-year EFS rate was 53.1% (± 9.4%) for the patients who received high-dose methotrexate and nelarabine. There was no difference in outcome for these two groups (HR, 0.66; 95% CI, 0.24–1.83; P = .423).[43]
Sources: AALL0434 (NCT00408005) · 43
For patients who achieve CR, measures of the rapidity of blast clearance and MRD determinations have important prognostic significance, particularly the following:
The percentage of morphologically detectable marrow blasts at 7 and 14 days after starting multiagent remission induction therapy has been correlated with relapse risk,[44] and has been used in the past by the COG to risk-stratify patients. However, in multivariate analyses, when end-induction MRD is included, these early marrow findings lose their prognostic significance.[45,46]
End-induction levels of submicroscopic MRD, assessed by multiparameter flow cytometry, polymerase chain reaction, or next-generation sequencing assays strongly correlates with long-term outcome.[45,47-50] Intensification of postinduction therapy for patients with high levels of end-induction MRD is a common component of most ALL treatment regimens. In a randomized trial conducted by the United Kingdom Acute Lymphoblastic Leukaemia (UKALL) group, augmented postinduction therapy was shown to improve outcome for standard-risk and intermediate-risk patients with high end-induction MRD.[51]
MRD levels earlier in induction (e.g., days 8 and 15) and at later postinduction time points (e.g., week 12 after starting therapy) have also been shown to have prognostic significance in both B-ALL and T-ALL.[45,46,49,52-55]
Nearly all patients with a positive end-of-induction MRD will become MRD negative at the end of 4 to 8 weeks of consolidation therapy. In a COG study, patients with high-risk B-ALL who had a positive end-of-induction MRD but a negative end-of-consolidation MRD had a significantly improved DFS compared with patients who were MRD positive at end of consolidation (5-year DFS rate, 79.5% vs. 39.5%).[46]
Sources: 46
For more information, see the Response to initial treatment section.
Sources: Response to initial treatment
For specific information about CNS therapy to prevent CNS relapse in children with newly diagnosed ALL, see the CNS-Directed Therapy for Childhood ALL section.
Standard Postinduction Treatment Options for Childhood ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Standard treatment options for consolidation/intensification and maintenance therapy (postinduction therapy) include the following:
Chemotherapy.
CNS-directed therapy is provided during premaintenance chemotherapy by all groups. Some protocols (COG, St. Jude Children's Research Hospital [SJCRH], and DFCI) provide ongoing intrathecal chemotherapy during maintenance, while others (Berlin-Frankfurt-Münster [BFM]) do not. For specific information about CNS therapy to prevent CNS relapse in children with acute lymphoblastic leukemia (ALL) who are receiving postinduction therapy, see the CNS-Directed Therapy for Childhood ALL section.
Consolidation/intensification therapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Once CR has been achieved, systemic treatment in conjunction with CNS-directed therapy follows. The intensity of the postinduction chemotherapy varies considerably depending on risk group assignment, but all patients receive some form of intensification after the achievement of CR and before beginning maintenance therapy.
The most commonly used intensification schema is the BFM backbone. This therapeutic backbone, first introduced by the BFM clinical trials group, includes the following:[1]
Sources: 1
An initial consolidation (referred to as induction IB) immediately after the initial induction phase. This phase includes intrathecal therapy, cyclophosphamide, low-dose cytarabine, and mercaptopurine.An interim maintenance phase, which includes intrathecal therapy and four doses of high-dose methotrexate (typically 5 g/m2) with leucovorin rescue.
Reinduction (or delayed intensification), which typically includes agents and schedules similar to those used during the induction and initial consolidation phases.
Maintenance, typically consisting of daily mercaptopurine (6-MP), weekly low-dose methotrexate, and sometimes, intermittent administration of vincristine and a corticosteroid, as well as continued intrathecal therapy.
This backbone has been adopted by many groups, including the COG. Variation of this backbone includes the following:
Intensification for higher-risk patients by including additional doses of vincristine and pegaspargase, as well as repeated interim maintenance and delayed intensification phases.[56,57]
The use of escalating doses of methotrexate (starting at a dose of 100 mg/m2) without leucovorin rescue instead of or in addition to high-dose methotrexate during interim maintenance phases.
Elimination or truncation of some of the phases for lower-risk patients to minimize acute and long-term toxicity.
Other clinical trial groups utilize a different therapeutic backbone during postinduction treatment phases, as follows:
DFCI: The DFCI ALL Consortium protocols include 30 weeks of pegaspargase therapy beginning at week 7 of therapy, given in conjunction with maintenance regimen (vincristine/dexamethasone pulses, weekly low-dose methotrexate, daily mercaptopurine).[3] These protocols also do not include a delayed intensification phase, but high-risk patients receive additional doses of doxorubicin (instead of low-dose methotrexate) during the first six months of postinduction therapy.
Sources: 3
NOPHO: The NOPHO also emphasizes the use of pegaspargase during consolidation and intensification. In the ALL2008 trial, all patients received five doses of pegaspargase given every other week after induction. Patients then received an additional ten doses at 2-week intervals or three doses at 6-week intervals. Both regimens produced equally excellent survival rates, with reduced toxicity in the three-dose regimen.[58]
Sources: 58
SJCRH: SJCRH follows a BFM backbone but augments the reinduction and maintenance phases for some patients by including intensified dosing of pegaspargase, frequent vincristine/corticosteroid pulses, and rotating drug pairs during maintenance (mercaptopurine/methotrexate, cyclophosphamide/cytarabine, dexamethasone/vincristine).[59]
Sources: 59
Standard-risk ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
In children with low- and standard-risk B-ALL, there has been an attempt to limit exposure to drugs such as anthracyclines and alkylating agents that may be associated with an increased risk of late toxic effects.[60-62] The COG regimen for standard-risk B-ALL postinduction therapy can be delivered in the outpatient setting and has multiple favorable characteristics, including low-intensity 4-week consolidation, limited anthracycline (75 mg/m2) and alkylator exposure (1 gm/m2), only two doses of pegaspargase, and interim maintenance phases consisting of escalating doses of methotrexate (without leucovorin rescue) rather than high-dose IV methotrexate.[63][Level of evidence B4]
Sources: 60 · 62 · 63 · Level of evidence B4
Favorable outcomes for standard-risk patients with B-ALL were also reported in trials that used a limited number of courses of intermediate-dose or high-dose methotrexate as consolidation followed by maintenance therapy (without a reinduction phase).[61,64,65] More specifically, a subset of patients with standard-risk B-ALL with favorable cytogenetics, no evidence of CNS or testicular disease at diagnosis, and rapid achievement of low levels of MRD, have been treated with exposure to no or low doses of anthracyclines and alkylating agents. The 5-year DFS rate was almost 99%, and the OS rate was 100%.[66] The DFCI ALL Consortium study used multiple doses of pegaspargase (30 weeks) as consolidation, without postinduction exposure to alkylating agents or anthracyclines.[67,68]
However, the prognostic impact of end-induction and/or consolidation MRD has influenced the treatment of patients originally diagnosed as NCI standard risk. Multiple studies have demonstrated that higher levels of end-induction MRD are associated with poorer prognosis.[45,47,48,69,70] Augmenting therapy has been shown to improve the outcome in standard-risk patients with elevated MRD levels at the end of induction.[51] Patients with NCI standard-risk B-ALL with high-risk features (including increased end-of-induction MRD levels as well as CNS2 status at diagnosis, and/or unfavorable genetics) are treated with more intensified therapy. For more information, see the Prognostic (risk) groups under clinical evaluation section.
Sources: 45 · 47 · 48 · 69 · 70 · 51 · Prognostic (risk) groups under clinical evaluation
Evidence (intensification for standard-risk B-ALL):
Clinical trials conducted in the 1980s and early 1990s demonstrated that the use of a delayed intensification phase improved outcome for children with standard-risk ALL treated with regimens using a BFM backbone.[71-73] The delayed intensification phase on such regimens, including those of the COG, consists of an 8-week phase of reinduction (including dexamethasone and an anthracycline) and reconsolidation containing cyclophosphamide, cytarabine, and 6-thioguanine given approximately 4 to 6 months after remission is achieved.[35,71,74]
The former Children's Cancer Group (CCG) study (CCG-1991/COG-1991) for standard-risk ALL used dexamethasone in a three-drug induction phase and tested the utility of a second delayed intensification phase. This study also compared escalating IV methotrexate (without leucovorin rescue) in conjunction with vincristine versus a standard maintenance combination with oral methotrexate given during two interim maintenance phases.[75][Level of evidence B1]
Sources: COG-1991 · 75 · Level of evidence B1
The former Children's Cancer Group (CCG) study (CCG-1991/COG-1991) for standard-risk ALL used dexamethasone in a three-drug induction phase and tested the utility of a second delayed intensification phase. This study also compared escalating IV methotrexate (without leucovorin rescue) in conjunction with vincristine versus a standard maintenance combination with oral methotrexate given during two interim maintenance phases.[75][Level of evidence B1]
A second delayed intensification phase provided no benefit in patients who were rapid early responders (M1 or M2 marrow by day 14 of induction).
Escalating IV methotrexate during the interim maintenance phases, compared with oral methotrexate during these phases, produced a significant improvement in EFS, which was because of a decreased incidence of isolated extramedullary relapses, particularly those involving the CNS.
Successful therapies for patients with standard-risk ALL that have decreased the use of drugs associated with long-term toxicities have focused on children with B-ALL, not T-ALL.[61,62,64,65,67,68] The COG, Dutch Children's Oncology Group (DCOG), DFCI, NOPHO, and other large cooperative groups have excluded patients with T-cell ALL from low and standard-risk therapies. Patients with NCI standard-risk features but a T-cell immunophenotype had an inferior EFS and OS compared with patients with NCI standard-risk B-ALL treated on the same regimens on CCG1952 and CCG1991.[76]
The COG AALL0331 (NCT00103285) study stratified intensity of therapy for NCI standard-risk patients on the basis of biology and early response. Rapid early response was defined as less than 5% bone marrow blasts by day 15 based on local morphological interpretation and an M1 bone marrow with MRD levels of less than 0.1% at day 29. Standard-risk low patients were those with favorable biology (ETV6::RUNX1 or high hyperdiploidy with triple trisomy), CNS1 status, and a rapid early response. Standard-risk average patients were those lacking favorable or unfavorable biology who also had a rapid early response. Standard-risk high patients were those with slow early response and/or CNS3 status, or KMT2A-rearranged patients with rapid early response. All patients received a three-drug prednisone-based induction (no anthracycline). Standard-risk average patients were randomly assigned to either intensified consolidation (augmented BFM) or standard consolidation. Standard-risk high patients were nonrandomly assigned to the full augmented BFM therapy used for NCI high-risk patients, including two delayed intensification phases.[77]
Sources: AALL0331 (NCT00103285) · 77
The COG AALL0331 (NCT00103285) study stratified intensity of therapy for NCI standard-risk patients on the basis of biology and early response. Rapid early response was defined as less than 5% bone marrow blasts by day 15 based on local morphological interpretation and an M1 bone marrow with MRD levels of less than 0.1% at day 29. Standard-risk low patients were those with favorable biology (ETV6::RUNX1 or high hyperdiploidy with triple trisomy), CNS1 status, and a rapid early response. Standard-risk average patients were those lacking favorable or unfavorable biology who also had a rapid early response. Standard-risk high patients were those with slow early response and/or CNS3 status, or KMT2A-rearranged patients with rapid early response. All patients received a three-drug prednisone-based induction (no anthracycline). Standard-risk average patients were randomly assigned to either intensified consolidation (augmented BFM) or standard consolidation. Standard-risk high patients were nonrandomly assigned to the full augmented BFM therapy used for NCI high-risk patients, including two delayed intensification phases.[77]
The 6-year EFS rate for all patients was 89%, and the OS rate was 96%.
For standard-risk low patients, this study evaluated the addition of four doses of pegaspargase (added in consolidation and interim maintenance phases) to standard therapy, which included two doses of pegaspargase (administered in induction and delayed intensification phases). Standard-risk low patients had highly favorable outcomes (6-year DFS and OS rates were 94.7% ± 0.6% and 98.7% ± 0.3%, respectively). The augmentation of standard-risk low therapy with additional pegaspargase did not improve outcomes.[63][Level of evidence B4]
Sources: 63 · Level of evidence B4
For standard-risk average patients, the augmented consolidation regimen did not improve rates of continuous complete remission (CCR) or OS. The 6-year rates of CCR and OS for the standard-risk average cohort were 88% to 89% and 95% to 96%, respectively.
Standard-risk average patients with end-induction MRD levels of 0.01% to <0.1% had an inferior outcome compared with those with MRD levels of <0.01% (6-year CCR rates, 77% vs. 91%, respectively). Augmented consolidation was not associated with a better outcome in standard-risk average patients with higher levels of MRD.
The standard-risk high cohort achieved a relatively favorable 6-year CCR rate of 86% and an OS rate of 93%.
In a randomized study conducted in the United Kingdom, children and young adults with ALL who lacked high-risk features (including adverse cytogenetics, and/or M3 marrow morphology at day 8 or day 15 of induction) were risk-stratified on the basis of MRD level at the end of induction (week 4) and at week 11 of therapy. Patients with undetectable MRD at week 4 (or with low MRD at week 4 and undetectable by week 11) were considered low risk, and were eligible to be randomly assigned to therapy with either one or two delayed intensification phases.[78][Level of evidence B1]
Sources: 78 · Level of evidence B1
In a randomized study conducted in the United Kingdom, children and young adults with ALL who lacked high-risk features (including adverse cytogenetics, and/or M3 marrow morphology at day 8 or day 15 of induction) were risk-stratified on the basis of MRD level at the end of induction (week 4) and at week 11 of therapy. Patients with undetectable MRD at week 4 (or with low MRD at week 4 and undetectable by week 11) were considered low risk, and were eligible to be randomly assigned to therapy with either one or two delayed intensification phases.[78][Level of evidence B1]
There was no significant difference in EFS between patients who received one and those who received two delayed intensification phases.
There was no significant difference in treatment-related deaths between the two arms; however, the second delayed intensification phase was associated with grade 3 or 4 toxic events in 17% of the 261 patients randomly assigned to that arm, and one patient experienced a treatment-related death during that phase.
In the AIEOP ALL-BFM-2000 (NCT00430118) trial, standard-risk patients (defined as those with undetectable MRD at days 33 and 78 and absence of high-risk cytogenetics) were randomly assigned to receive treatment with a single delayed-intensification phase of either standard intensity or reduced intensity (shorter duration, with reduced total dosages of dexamethasone, vincristine, doxorubicin, and cyclophosphamide).[79]
Sources: ALL-BFM-2000 (NCT00430118) · 79
In the AIEOP ALL-BFM-2000 (NCT00430118) trial, standard-risk patients (defined as those with undetectable MRD at days 33 and 78 and absence of high-risk cytogenetics) were randomly assigned to receive treatment with a single delayed-intensification phase of either standard intensity or reduced intensity (shorter duration, with reduced total dosages of dexamethasone, vincristine, doxorubicin, and cyclophosphamide).[79]
Reduced-intensity delayed intensification was associated with an inferior 8-year DFS rate (89% vs. 92%, P = .04), resulting from an increased risk of relapse.
In a subset analysis, for patients with the ETV6::RUNX1 fusion, no difference in outcome between the two treatment arms was observed (8-year DFS rate, approximately 94% for both arms).
The Malaysia-Singapore ALL MS2010 trial for patients with favorable-risk B-ALL evaluated a deintensified, modified BFM regimen. This regimen omitted anthracyclines, had fewer doses of high-dose methotrexate, and fewer doses of low-dose cytarabine.[80]
Sources: 80
The Malaysia-Singapore ALL MS2010 trial for patients with favorable-risk B-ALL evaluated a deintensified, modified BFM regimen. This regimen omitted anthracyclines, had fewer doses of high-dose methotrexate, and fewer doses of low-dose cytarabine.[80]
The long-term EFS in this trial (6-year EFS rate, 96.5%) was noninferior, compared with the predecessor trial conducted by the same group.
This regimen was also less toxic, with significantly decreased rates of bacteremia and septic shock/intensive care unit admissions.
Patients who are standard or intermediate risk at diagnosis, but have high levels of end-induction MRD, have been shown to have a poorer prognosis and should be treated as high-risk patients. The UKALL2003 (NCT00222612) trial used augmented postinduction therapy (extra doses of pegaspargase and vincristine and an escalated-dose of IV methotrexate without leucovorin rescue) to treat standard- or intermediate-risk patients with high levels of end-induction MRD.[51][Level of evidence B1]
Sources: UKALL2003 (NCT00222612) · 51 · Level of evidence B1
Patients who are standard or intermediate risk at diagnosis, but have high levels of end-induction MRD, have been shown to have a poorer prognosis and should be treated as high-risk patients. The UKALL2003 (NCT00222612) trial used augmented postinduction therapy (extra doses of pegaspargase and vincristine and an escalated-dose of IV methotrexate without leucovorin rescue) to treat standard- or intermediate-risk patients with high levels of end-induction MRD.[51][Level of evidence B1]
Augmented postinduction therapy resulted in an increased EFS that was comparable to that of patients with low levels of end-induction MRD.
High-risk ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
In high-risk patients, a number of different approaches have been used with comparable efficacy.[67,81]; [74][Level of evidence B4] Treatment for high-risk patients is generally more intensive than that for standard-risk patients and typically includes higher cumulative doses of multiple agents, including anthracyclines and/or alkylating agents. Higher doses of these agents increase the risk of both short-term and long-term toxicities, and many clinical trials have focused on reducing the side effects of these intensified regimens.
Sources: 67 · 81 · 74 · Level of evidence B4
Evidence (intensification for high-risk ALL):
The former CCG developed an augmented BFM treatment regimen that included a second interim maintenance and delayed intensification phase. This regimen featured repeated courses of escalating-dose IV methotrexate (without leucovorin rescue) given with vincristine and pegaspargase during interim maintenance and additional vincristine and pegaspargase pulses during initial consolidation and delayed intensification. In the CCG-1882 trial, NCI high-risk patients with slow early response (M3 marrow on day 7 of induction) were randomly assigned to receive either standard- or augmented-BFM therapy.[56]
Sources: 56
The former CCG developed an augmented BFM treatment regimen that included a second interim maintenance and delayed intensification phase. This regimen featured repeated courses of escalating-dose IV methotrexate (without leucovorin rescue) given with vincristine and pegaspargase during interim maintenance and additional vincristine and pegaspargase pulses during initial consolidation and delayed intensification. In the CCG-1882 trial, NCI high-risk patients with slow early response (M3 marrow on day 7 of induction) were randomly assigned to receive either standard- or augmented-BFM therapy.[56]
The augmented-therapy regimen in the CCG-1882 trial produced significantly better EFS and OS rates (75% and 78%) than did the standard CCG modified-BFM therapy (55% and 66.7%).
There was a significantly higher incidence of osteonecrosis in patients older than 10 years who received the augmented therapy (which included two 21-day postinduction dexamethasone courses), compared with those who were treated on the standard arm (one 21-day postinduction dexamethasone course).[82]
Sources: 82
In an Italian study, investigators showed that two applications of delayed intensification therapy (protocol II) significantly improved outcome for patients with a poor response to a prednisone prophase.[83]
Sources: 83
The CCG-1961 study used a 2 × 2 factorial design to compare both standard- versus augmented-intensity therapies and therapies of standard duration (one interim maintenance and delayed intensification phase) versus increased duration (two interim maintenance and delayed intensification phases) among NCI high-risk patients with a rapid early response. This trial also tested whether continuous versus alternate-week dexamethasone during delayed intensification phases affected rates of osteonecrosis.
Sources: CCG-1961
The CCG-1961 study used a 2 × 2 factorial design to compare both standard- versus augmented-intensity therapies and therapies of standard duration (one interim maintenance and delayed intensification phase) versus increased duration (two interim maintenance and delayed intensification phases) among NCI high-risk patients with a rapid early response. This trial also tested whether continuous versus alternate-week dexamethasone during delayed intensification phases affected rates of osteonecrosis.
Augmented therapy was associated with an improvement in EFS. There was no EFS benefit associated with the administration of the second interim maintenance and delayed intensification phases.[57,84][Level of evidence A1]
Sources: 57 · 84 · Level of evidence A1
The cumulative incidence of osteonecrosis at 5 years was 9.9% for patients aged 10 to 15 years and 20.0% for patients aged 16 to 21 years, compared with 1.0% for patients aged 1 to 9 years (P = .0001). For patients aged 10 to 21 years, alternate-week dosing of dexamethasone during delayed intensification phases was associated with a significantly lower cumulative incidence of osteonecrosis, compared with continuous dosing (8.7% vs. 17.0%, P = .0005).[85][Level of evidence A3]
Sources: 85 · Level of evidence A3
In the UKALL2003 (NCT00222612) trial, patients with high end-induction MRD (>0.01%) and/or high-risk cytogenetics were randomly assigned to receive either a standard-intensity or an augmented BFM chemotherapy backbone.[86]
Sources: UKALL2003 (NCT00222612) · 86
In the UKALL2003 (NCT00222612) trial, patients with high end-induction MRD (>0.01%) and/or high-risk cytogenetics were randomly assigned to receive either a standard-intensity or an augmented BFM chemotherapy backbone.[86]
The 10-year EFS rate was 87.1% for patients assigned to the augmented chemotherapy backbone, compared with 82.1% for those assigned the standard-intensity chemotherapy backbone (P = .09).
Patients with high-risk cytogenetics had a significantly lower risk of relapse when treated with augmented therapy (10-year relapse rate, 22.1% vs. 52.4% with standard-intensity therapy; P = .016).
In the COG AALL0232 (NCT00075725) study (2004–2011), patients with high-risk B-ALL received an augmented BFM backbone with one interim maintenance and delayed intensification phase. Only patients with end-induction MRD greater than 0.1% or M2/M3 marrow at day 15 received two interim maintenance/delayed intensification phases. Patients were randomly assigned to receive either high-dose methotrexate or escalating dose IV methotrexate (Capizzi methotrexate) plus pegaspargase during the interim maintenance phase (the first phase only for those receiving two of these phases).[8,46]
Sources: AALL0232 (NCT00075725) · 8 · 46
In the COG AALL0232 (NCT00075725) study (2004–2011), patients with high-risk B-ALL received an augmented BFM backbone with one interim maintenance and delayed intensification phase. Only patients with end-induction MRD greater than 0.1% or M2/M3 marrow at day 15 received two interim maintenance/delayed intensification phases. Patients were randomly assigned to receive either high-dose methotrexate or escalating dose IV methotrexate (Capizzi methotrexate) plus pegaspargase during the interim maintenance phase (the first phase only for those receiving two of these phases).[8,46]
The methotrexate randomization was terminated early when planned interim monitoring indicated that high-dose methotrexate was associated with superior outcome. The 5-year EFS rate of patients randomly assigned to high-dose methotrexate was 79.6%, compared with 75% for those randomly assigned to the Capizzi methotrexate arm. High-dose methotrexate was also associated with a superior 5-year OS (P = .025).[8]
Sources: 8
Patients with MRD less than 0.01% at end of induction had a 5-year EFS rate of 87%, compared with 74% for those with MRD 0.01% to 0.1%. Those with MRD levels greater than 0.1% fared worse.[46]
Sources: 46
High-dose methotrexate was associated with a superior EFS rate in patients with end-induction MRD greater than 0.01% (high-dose methotrexate, 68%; Capizzi methotrexate, 58%; P = .008).[46]
Sources: 46
Because treatment for high-risk ALL involves more intensive therapy, leading to a higher risk of acute and long-term toxicities, a number of clinical trials have tested interventions to prevent side effects without adversely impacting EFS. Interventions that have been investigated include the use of the cardioprotectant dexrazoxane to prevent anthracycline-related cardiac toxic effects and alternative scheduling of corticosteroids to reduce the risk of osteonecrosis.
Evidence (cardioprotective effect of dexrazoxane):
In a DFCI ALL Consortium trial, children with high-risk ALL were randomly assigned to receive doxorubicin alone (30 mg/m2/dose to a cumulative dose of 300 mg/m2) or with dexrazoxane during the induction and intensification phases of multiagent chemotherapy.[87,88]
In a DFCI ALL Consortium trial, children with high-risk ALL were randomly assigned to receive doxorubicin alone (30 mg/m2/dose to a cumulative dose of 300 mg/m2) or with dexrazoxane during the induction and intensification phases of multiagent chemotherapy.[87,88]
The use of the cardioprotectant dexrazoxane before doxorubicin resulted in better left ventricular fractional shortening and improved end-systolic dimension Z-scores without any adverse effect on EFS or increase in second malignancy risk, compared with the use of doxorubicin alone 5 years posttreatment.
A greater long-term protective effect was noted in girls than in boys.
On the POG-9404 trial, patients with T-ALL were randomly assigned to receive dexrazoxane or not before each dose of doxorubicin (cumulative dose 360 mg/m2).[89]
On the POG-9404 trial, patients with T-ALL were randomly assigned to receive dexrazoxane or not before each dose of doxorubicin (cumulative dose 360 mg/m2).[89]
There was no difference in EFS between patients with T-ALL who were treated with dexrazoxane and patients who were not treated with dexrazoxane (cumulative doxorubicin dose, 360 mg/m2).
Three years after initial diagnosis, left ventricular shortening fraction and left ventricular wall thickness were both significantly worse in patients who received doxorubicin alone than in patients who received dexrazoxane, indicating that dexrazoxane was cardioprotective. The frequency of grade 3 and 4 toxicities that occurred during therapy was similar between the randomized groups, and there was no difference in cumulative incidence of second malignant neoplasms.
Evidence (reducing risk of osteonecrosis):
In the CCG-1961 study, alternate-week dosing of dexamethasone during delayed intensification was studied with the goal of reducing the frequency of osteonecrosis.[85] Patients with high-risk B-ALL and a rapid early morphological response to induction therapy were randomly assigned to receive either one or two delayed intensification phases. Patients randomly assigned to one delayed intensification phase received daily dosing of dexamethasone (21 consecutive days), while those randomly assigned to two delayed intensification phases received alternate-week dosing of dexamethasone (days 0–6 and 14–21) during each delayed intensification phase.
Sources: 85
In the CCG-1961 study, alternate-week dosing of dexamethasone during delayed intensification was studied with the goal of reducing the frequency of osteonecrosis.[85] Patients with high-risk B-ALL and a rapid early morphological response to induction therapy were randomly assigned to receive either one or two delayed intensification phases. Patients randomly assigned to one delayed intensification phase received daily dosing of dexamethasone (21 consecutive days), while those randomly assigned to two delayed intensification phases received alternate-week dosing of dexamethasone (days 0–6 and 14–21) during each delayed intensification phase.
For patients aged 10 years or older at diagnosis, those who received two delayed intensification phases (alternate-week dosing of dexamethasone) had a significantly lower risk of symptomatic osteonecrosis (5-year cumulative incidence of 8.7%, compared with 17% for patients receiving one delayed intensification phase with continuous dexamethasone dosing; P = .001).
The greatest impact was seen in females aged 16 to 21 years, who showed the highest incidence of osteonecrosis with standard therapy containing continuous dexamethasone; the incidence of osteonecrosis with alternative-week dexamethasone was 5.6%, compared with 57.6% for those receiving continuous dosing.
For more information, see the Osteonecrosis section.
Sources: Osteonecrosis
Very high-risk ALL
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Approximately 10% to 20% of patients with ALL are classified as very high risk, including the following:[74,90]
Infants younger than 1 year, especially if there is a KMT2A gene rearrangement present. For more information about infants with ALL, see the Infants With ALL section.
Sources: Infants With ALL
Patients with adverse cytogenetic abnormalities, including BCR::ABL1, TCF3::HLF, KMT2A gene rearrangements, and low hypodiploidy (<44 chromosomes).
Patients who achieve CR but have a slow early response to initial therapy, including those with a high absolute blast count after a 7-day steroid prophase, and patients with high MRD levels at the end of induction (week 4) or later time points (e.g., week 12).
Patients who have morphologically persistent disease after the first 4 weeks of therapy (induction failure), even if they later achieve CR.
Patients with very high-risk features have been treated with multiple cycles of intensive chemotherapy during the consolidation phase (usually in addition to the typical BFM backbone intensification phases). These additional cycles often include agents not typically used in frontline ALL regimens for standard-risk and high-risk patients, such as high-dose cytarabine, ifosfamide, and etoposide.[74] However, even with this intensified approach, reported long-term EFS rates range from 30% to 50% for some of these very high-risk subsets.[39,74] The DCOG reported the outcomes of 107 patients with very high-risk features who were treated with three to six intensive chemotherapy blocks in two consecutive trials. Sixty of these patients received an allogeneic HSCT in first CR. The 5-year EFS rate was 73%, and the OS rate was 79% for all patients. With this intensified treatment approach, the cumulative incidence of treatment-related mortality was 12.3%, which was similar to the cumulative incidence of relapse, at 13%.[91]
On some clinical trials, very high-risk patients have also been considered candidates for allogeneic HSCT in first CR.[39,92-94] However, there are limited data regarding the outcome of very high-risk patients treated with allogeneic HSCT in first CR. Controversy exists regarding which subpopulations could potentially benefit from HSCT.
Evidence (allogeneic HSCT in first remission for very high-risk patients):
In a European cooperative group study conducted between 1995 and 2000, very high-risk patients were defined as one of the following: morphologically persistent disease after a four-drug induction, BCR::ABL1 or KMT2A::AFF1 fusions, or poor response to prednisone prophase in patients with either T-cell phenotype or presenting white blood cells (WBC) >100,000/μL. These patients were assigned to receive either an allogeneic HSCT in first CR (based on the availability of a human lymphocyte antigen–matched related donor) or intensive chemotherapy.[39]
Sources: 39
In a European cooperative group study conducted between 1995 and 2000, very high-risk patients were defined as one of the following: morphologically persistent disease after a four-drug induction, BCR::ABL1 or KMT2A::AFF1 fusions, or poor response to prednisone prophase in patients with either T-cell phenotype or presenting white blood cells (WBC) >100,000/μL. These patients were assigned to receive either an allogeneic HSCT in first CR (based on the availability of a human lymphocyte antigen–matched related donor) or intensive chemotherapy.[39]
Using an intent-to-treat analysis, patients assigned to allogeneic HSCT (on the basis of donor availability) had a superior 5-year DFS rate compared with patients assigned to intensive chemotherapy (57% ± 7% for transplant vs. 41% ± 3% for chemotherapy, P = .02).
There was no significant difference in OS rates (56% ± 6% for transplant vs. 50% ± 3% for chemotherapy; P = .12).
For patients with T-ALL and a poor response to prednisone prophase, both DFS and OS rates were significantly better with allogeneic HSCT.[92]
Sources: 92
In a large retrospective series of patients with initial induction failure, the 10-year OS rate for patients with persistent leukemia was 32%.[41]
Sources: 41
In a large retrospective series of patients with initial induction failure, the 10-year OS rate for patients with persistent leukemia was 32%.[41]
A trend for superior outcome with allogeneic HSCT, compared with chemotherapy alone, was observed in patients with T-cell phenotype (any age) and with B-ALL who were older than 6 years.
Patients with B-ALL who were aged 1 to 5 years at diagnosis and did not have any adverse cytogenetic abnormalities (KMT2A rearrangement, BCR::ABL1) had a relatively favorable prognosis, without any advantage in outcome with the utilization of HSCT compared with chemotherapy alone.
The AIEOP ALL-BFM-2000 (NCT00430118) study (2000–2006) classified patients as high risk if they met any of the following criteria: poor response to prednisone prophase, failure to achieve CR at the end of the first month of treatment, high MRD levels after induction IB (day 78 of therapy), and KMT2A::AFF1 fusion. These patients were allocated to allogeneic HSCT in first CR per protocol on the basis of donor availability and investigator preference.[95][Level of evidence B4]
Sources: ALL-BFM-2000 (NCT00430118) · 95 · Level of evidence B4
The AIEOP ALL-BFM-2000 (NCT00430118) study (2000–2006) classified patients as high risk if they met any of the following criteria: poor response to prednisone prophase, failure to achieve CR at the end of the first month of treatment, high MRD levels after induction IB (day 78 of therapy), and KMT2A::AFF1 fusion. These patients were allocated to allogeneic HSCT in first CR per protocol on the basis of donor availability and investigator preference.[95][Level of evidence B4]
The overall 5-year EFS rate of patients meeting high-risk criteria was 58.9%.
The 5-year EFS rate was 74% for patients whose only high-risk feature was prednisone-poor response. There was no significant difference in DFS (P = .31) or OS (P = .91) when comparing HSCT and chemotherapy for patients with poor prednisone response in whom HSCT was allowed per protocol (those with T-ALL and/or WBC ≥100,000/mm3).
All other high-risk patients (i.e., those with initial induction failure, high day 78 MRD and/or KMT2A::AFF1 fusion) had EFS rates less than 50%. For these patients, there was no statistically significant difference in DFS between those who received HSCT (n = 66) and those who received chemotherapy only (n = 88), after adjusting for waiting time to HSCT (5.7 months).
On the NOPHO ALL2008 (NCT00819351) protocol, patients were allocated to HSCT in first CR if they had MRD levels of 5% or greater at the end of induction or MRD levels of 0.1% or greater at end of consolidation. All patients allocated to HSCT received at least three blocks of intensive chemotherapy before HSCT to reduce levels of MRD.[96]
Sources: ALL2008 (NCT00819351) · 96
On the NOPHO ALL2008 (NCT00819351) protocol, patients were allocated to HSCT in first CR if they had MRD levels of 5% or greater at the end of induction or MRD levels of 0.1% or greater at end of consolidation. All patients allocated to HSCT received at least three blocks of intensive chemotherapy before HSCT to reduce levels of MRD.[96]
In the intent-to-treat analysis of 69 patients who met HSCT criteria (10 of whom did not undergo HSCT), the 5-year DFS rate was 78%.
Comparing the patients in this cohort who did and did not receive HSCT, receipt of HSCT was not significantly associated with survival (HR, 1.4; P = .69).
For patients who underwent HSCT, superior outcomes (better DFS and lower cumulative incidence of relapse) were observed in patients who had nondetectable MRD before HSCT.
In the DCOG ALL-10 and ALL-11 trials, patients with very high-risk features received an intensified treatment regimen that included three high-dose chemotherapy blocks after consolidation. After the three blocks, 60 patients received an allogeneic HSCT, and 22 patients continued with a chemotherapy-only approach, which included three additional high-dose blocks. In these trials, very high-risk disease was defined by any of the following features: morphologically detectable disease at end of induction, high end-consolidation MRD (time point 2), t(4;11), or poor response to a prednisone prophase.[91]
Sources: 91
In the DCOG ALL-10 and ALL-11 trials, patients with very high-risk features received an intensified treatment regimen that included three high-dose chemotherapy blocks after consolidation. After the three blocks, 60 patients received an allogeneic HSCT, and 22 patients continued with a chemotherapy-only approach, which included three additional high-dose blocks. In these trials, very high-risk disease was defined by any of the following features: morphologically detectable disease at end of induction, high end-consolidation MRD (time point 2), t(4;11), or poor response to a prednisone prophase.[91]
The 5-year EFS rate was 72.8% for all patients.
In a landmark analysis of EFS from the end of the third high-dose chemotherapy block, no difference was observed in the outcomes of patients who received HSCT versus those who received chemotherapy only.
Two retrospective analyses investigated the role of HSCT in first CR for patients with hypodiploid ALL. The studies showed no clear evidence that HSCT improved outcomes when 1) transplanting all patients with hypodiploid ALL, or 2) transplanting hypodiploid patients deemed at high risk on the basis of high MRD after induction. The studies did not examine the strategy of HSCT for persistent MRD after consolidation, nor did they analyze the status of MRD at the time of HSCT.
Two retrospective analyses investigated the role of HSCT in first CR for patients with hypodiploid ALL. The studies showed no clear evidence that HSCT improved outcomes when 1) transplanting all patients with hypodiploid ALL, or 2) transplanting hypodiploid patients deemed at high risk on the basis of high MRD after induction. The studies did not examine the strategy of HSCT for persistent MRD after consolidation, nor did they analyze the status of MRD at the time of HSCT.
In a study of 306 hypodiploid patients from 16 ALL cooperative groups treated between 1997 and 2013, a subgroup of 228 patients (42 who underwent HSCT) with 44 or fewer chromosomes who achieved remission were analyzed.[97][Level of evidence C2]
Sources: 97 · Level of evidence C2
Two retrospective analyses investigated the role of HSCT in first CR for patients with hypodiploid ALL. The studies showed no clear evidence that HSCT improved outcomes when 1) transplanting all patients with hypodiploid ALL, or 2) transplanting hypodiploid patients deemed at high risk on the basis of high MRD after induction. The studies did not examine the strategy of HSCT for persistent MRD after consolidation, nor did they analyze the status of MRD at the time of HSCT. › In a study of 306 hypodiploid patients from 16 ALL cooperative groups treated between 1997 and 2013, a subgroup of 228 patients (42 who underwent HSCT) with 44 or fewer chromosomes who achieved remission were analyzed.[97][Level of evidence C2]
Favorable prognostic factors included a chromosome number of 44 (compared with 43 or fewer), MRD less than 0.01% after induction, and treatment on an MRD-stratified protocol that intensified therapy for patients with higher MRD after induction.
After correction for median time to transplant, patients with low MRD who underwent HSCT had a DFS rate of 73.6%, compared with a DFS rate of 70% for those treated with chemotherapy alone (P = .81). Patients with higher MRD after induction who underwent HSCT had a DFS rate of 55.9%, compared with a DFS rate of 40.3% for those treated with chemotherapy (P = .29).
Two retrospective analyses investigated the role of HSCT in first CR for patients with hypodiploid ALL. The studies showed no clear evidence that HSCT improved outcomes when 1) transplanting all patients with hypodiploid ALL, or 2) transplanting hypodiploid patients deemed at high risk on the basis of high MRD after induction. The studies did not examine the strategy of HSCT for persistent MRD after consolidation, nor did they analyze the status of MRD at the time of HSCT.
The COG published an analysis of 113 evaluable patients with hypodiploid ALL who were treated between 2003 and 2011; 61 of those patients underwent HSCT in first CR.[98][Level of evidence C1]
Sources: 98 · Level of evidence C1
Two retrospective analyses investigated the role of HSCT in first CR for patients with hypodiploid ALL. The studies showed no clear evidence that HSCT improved outcomes when 1) transplanting all patients with hypodiploid ALL, or 2) transplanting hypodiploid patients deemed at high risk on the basis of high MRD after induction. The studies did not examine the strategy of HSCT for persistent MRD after consolidation, nor did they analyze the status of MRD at the time of HSCT. › The COG published an analysis of 113 evaluable patients with hypodiploid ALL who were treated between 2003 and 2011; 61 of those patients underwent HSCT in first CR.[98][Level of evidence C1]
The 5-year EFS rate was 57.4% for patients who underwent HSCT and 47.8% for patients in the chemotherapy cohorts (P = .49). The OS rate was 66.2% for patients who underwent HSCT and 53.8% for patients in the chemotherapy cohorts (P = .34).
Patients with high MRD after induction (≥0.01%) had a very poor EFS rate of 26.7% at 5 years, with no difference between the patients who received HSCT and the patients who received chemotherapy.
Maintenance therapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Backbone of maintenance therapy
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
The backbone of maintenance therapy in most protocols includes daily oral mercaptopurine and weekly oral or parenteral methotrexate. On many protocols, intrathecal chemotherapy for CNS sanctuary therapy is continued during maintenance therapy. Also, vincristine/steroid pulses during maintenance are used by some groups but not others (see below). It is imperative to carefully monitor children on maintenance therapy for both drug-related toxicity and for compliance with the oral chemotherapy agents used during maintenance therapy.[99] A protocol conducted by the COG suggested there are significant differences in compliance with oral mercaptopurine regimens among various racial and socioeconomic groups and that level of adherence impacts relapse risk.[99,100]
In the past, clinical practice generally called for the administration of oral mercaptopurine in the evening, on the basis of evidence from older studies that this practice may improve EFS.[101] However, in a study conducted by the NOPHO group, in which details of oral intake were prospectively captured, timing of mercaptopurine administration (nighttime vs. other times of day) was not of prognostic significance.[102] In a COG study, taking mercaptopurine at varying times of day rather than consistently at nighttime was associated with higher rates of nonadherence. However, among adherent patients (i.e., those who took >95% of prescribed doses), there was no association between timing of mercaptopurine ingestion and relapse risk.[103]
Some patients may develop severe hematologic toxicity when receiving conventional dosages of mercaptopurine because of an inherited deficiency (homozygous mutant) of thiopurine S-methyltransferase, an enzyme that inactivates mercaptopurine.[104,105] These patients are able to tolerate mercaptopurine only in much lower dosages than those conventionally used.[104,105] Patients who are heterozygous for the variant generally tolerate mercaptopurine without serious toxicity, but they do require more frequent dose reductions for hematologic toxicity than do patients who are homozygous for the normal allele.[104] Polymorphisms of the NUDT15 gene, observed most frequently in East Asian and Hispanic patients, have also been linked to extreme sensitivity to the myelosuppressive effects of mercaptopurine.[106-108]
Evidence (maintenance therapy):
A meta-analysis of randomized trials compared thiopurines and found the following:
A meta-analysis of randomized trials compared thiopurines and found the following:
Thioguanine did not improve the overall EFS, although particular subgroups may benefit from its use.[109]
Sources: 109
The use of continuous thioguanine instead of mercaptopurine during the maintenance phase is associated with an increased risk of hepatic complications, including veno-occlusive disease (sinusoidal obstruction syndrome) and portal hypertension.[110-114]
Because of the increased toxicity of thioguanine, mercaptopurine remains the standard drug of choice.
In the COG AALL0932 (NCT01190930) trial, NCI standard-risk patients with average-risk features were randomly assigned to receive weekly oral methotrexate during maintenance at one of two starting doses: 20 mg/m2 (standard) or 40 mg/m2 (investigational).[115][Level of evidence A1]
Sources: AALL0932 (NCT01190930) · 115 · Level of evidence A1
In the COG AALL0932 (NCT01190930) trial, NCI standard-risk patients with average-risk features were randomly assigned to receive weekly oral methotrexate during maintenance at one of two starting doses: 20 mg/m2 (standard) or 40 mg/m2 (investigational).[115][Level of evidence A1]
There was no significant difference in 5-year DFS from the start of maintenance therapy between the two treatment arms (5-year DFS rate, 95.1% for patients who received the standard dose vs. 94.2% for patients who received the investigational dose; P = .92), indicating no advantage for the higher dose of oral methotrexate.
An intensified maintenance regimen, consisting of rotating pairs of agents, including cyclophosphamide and epipodophyllotoxins along with more standard maintenance agents, has been evaluated in several clinical trials conducted by SJCRH and other groups.[2]
Sources: 2
An intensified maintenance regimen, consisting of rotating pairs of agents, including cyclophosphamide and epipodophyllotoxins along with more standard maintenance agents, has been evaluated in several clinical trials conducted by SJCRH and other groups.[2]
The intensified maintenance with rotating pairs of agents was associated with more episodes of febrile neutropenia [116] and a higher risk of secondary acute myelogenous leukemia,[117,118] especially when epipodophyllotoxins were included.[116]On the basis of these findings, SJCRH modified the agents used in the rotating pair schedule during the maintenance phase. On the Total XV study, standard-risk and high-risk patients received three rotating pairs (mercaptopurine plus methotrexate, cyclophosphamide plus cytarabine, and dexamethasone plus vincristine) throughout this treatment phase. Low-risk patients received more standard maintenance (without cyclophosphamide and cytarabine).[59]
A randomized study from Argentina demonstrated no benefit from this intensified approach compared with a more standard maintenance regimen for patients who receive induction and consolidation phases based on a BFM backbone.[116]
Sources: 116
Vincristine/corticosteroid pulses
Source: Childhood Acute Lymphoblastic Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 21, 2025. Captured 2026-09-09.
Pulses of vincristine and corticosteroid are often added to the standard maintenance backbone, although the benefit of these pulses within the context of contemporary multiagent chemotherapy regimens remains controversial.
Evidence (vincristine/corticosteroid pulses):
A CCG randomized trial conducted in the 1980s demonstrated improved outcome in patients who received monthly vincristine/prednisone pulses.[119]
Sources: 119
A meta-analysis combining data from six clinical trials from the same treatment era showed an EFS advantage for vincristine/prednisone pulses.[120,121] However, overall EFS from these trials was lower than is observed with more contemporary regimens.
A systematic review of the impact of vincristine plus steroid pulses from more recent clinical trials raised the question of whether such pulses are of value in current ALL treatment, which includes more intensive early therapy and risk stratification incorporating early response (MRD) and biological factors.[121]
Sources: 121
In a multicenter randomized trial in children with intermediate-risk ALL being treated on a BFM regimen, there was no benefit associated with the addition of six pulses of vincristine/dexamethasone during the continuation phase, although the pulses were administered less frequently than in other trials in which a benefit had been demonstrated.[122]
Sources: 122
A small multicenter trial of average-risk patients demonstrated superior EFS in patients receiving vincristine plus corticosteroid pulses. In this study, there was no difference in outcome based on type of steroid (prednisone vs. dexamethasone).[123][Level of evidence A1]
Sources: 123 · Level of evidence A1
In the COG AALL0932 (NCT01190930) trial, standard-risk patients were randomly assigned during maintenance to receive vincristine/dexamethasone pulses every 4 weeks or every 12 weeks.[115][Level of evidence A1]
Sources: AALL0932 (NCT01190930) · 115 · Level of evidence A1
In the COG AALL0932 (NCT01190930) trial, standard-risk patients were randomly assigned during maintenance to receive vincristine/dexamethasone pulses every 4 weeks or every 12 weeks.[115][Level of evidence A1]
For the randomly assigned standard-risk patients, the 5-year DFS rate from the start of maintenance was 94.6%. There was no significant difference between the every-4-week group and the every-12-week group.
The Chinese Children’s Cancer Group conducted a randomized noninferiority trial to determine whether vincristine/dexamethasone pulses could be omitted during the second year of maintenance therapy. One year after the initiation of therapy, 5,054 patients with non-BCR::ABL1 fusion–positive ALL (B-ALL and T-ALL, aged 0–18 years) were randomly assigned to receive either vincristine/dexamethasone pulses every 8 weeks (seven pulses total) or no pulses during the second year of maintenance chemotherapy. Noninferiority was defined by calculating the one-sided 95% upper confidence bound of the difference in EFS probability between arms to ensure that an EFS decrement of 5% or more was ruled out.[124]
Sources: 124
The Chinese Children’s Cancer Group conducted a randomized noninferiority trial to determine whether vincristine/dexamethasone pulses could be omitted during the second year of maintenance therapy. One year after the initiation of therapy, 5,054 patients with non-BCR::ABL1 fusion–positive ALL (B-ALL and T-ALL, aged 0–18 years) were randomly assigned to receive either vincristine/dexamethasone pulses every 8 weeks (seven pulses total) or no pulses during the second year of maintenance chemotherapy. Noninferiority was defined by calculating the one-sided 95% upper confidence bound of the difference in EFS probability between arms to ensure that an EFS decrement of 5% or more was ruled out.[124]
For low-risk patients (NCI standard-risk B-ALL with high hyperdiploidy or ETV6::RUNX1 and low end-induction MRD), the EFS difference between arms met the protocol definition of noninferiority, indicating that omission of vincristine/dexamethasone pulses during the second year of maintenance did not result in a decrement of EFS that was greater than 5%.
For intermediate-risk and high-risk patients, the difference in 5-year EFS between arms did not meet the protocol definition of noninferiority (the 95% upper confidence bound for the difference was 0.055, which exceeded the preset noninferiority margin of 0.05); therefore, it could not be concluded that vincristine/dexamethasone pulses could be omitted in these patients without resulting in an EFS decrement exceeding 5%.
The Japanese Childhood Cancer Group ALL-B12 trial included 1,107 patients with standard-risk B-ALL (i.e., NCI standard risk, good response to prednisone, day 15 marrow M1/2, end-of-induction marrow M1 without high-risk features [CNS3, KMT2A rearranged, TCF3::HLF fusion, TCF4::HLF fusion, hypodiploid]). Patients were randomly assigned to receive maintenance therapy, either with or without every 4-week vincristine/dexamethasone pulses.[125]
Sources: 125
The Japanese Childhood Cancer Group ALL-B12 trial included 1,107 patients with standard-risk B-ALL (i.e., NCI standard risk, good response to prednisone, day 15 marrow M1/2, end-of-induction marrow M1 without high-risk features [CNS3, KMT2A rearranged, TCF3::HLF fusion, TCF4::HLF fusion, hypodiploid]). Patients were randomly assigned to receive maintenance therapy, either with or without every 4-week vincristine/dexamethasone pulses.[125]
There was no difference in EFS or adverse events between the two treatment groups.
A systematic review and meta-analysis evaluated the effect of reducing vincristine/steroid pulses on EFS, OS, and toxicity in patients with B-ALL. Twenty-five publications that included more than 12,000 patients were examined.[126]
Sources: 126
A systematic review and meta-analysis evaluated the effect of reducing vincristine/steroid pulses on EFS, OS, and toxicity in patients with B-ALL. Twenty-five publications that included more than 12,000 patients were examined.[126]
This study demonstrated that the benefit of these pulses noted in historical trials was not seen in contemporary trials.
However, there was an increased risk of grade 3+ nonhepatic toxicity in the high-pulse frequency group.
The authors concluded that decreasing or removing pulses likely does not affect survival or risk of relapse, but it is associated with reduced toxicity.
For regimens that include vincristine/steroid pulses, a number of studies have addressed which steroid (dexamethasone or prednisone) should be used. From these studies, it appears that dexamethasone is associated with superior EFS, but also may lead to a greater frequency of steroid-associated complications, including bone toxicity and infections, especially in older children and adolescents.[6,7,24,71,127] Compared with prednisone, dexamethasone has also been associated with a higher frequency of behavioral problems.[7] In a randomized study of 50 patients aged 3 to 16 years who received maintenance chemotherapy, concurrent administration of hydrocortisone (at physiological dosing) during dexamethasone pulses reduced the frequency of behavioral difficulties, emotional lability, and sleep disturbances.[128]
Evidence (dexamethasone vs. prednisone):
In a CCG study, dexamethasone was compared with prednisone during the induction and maintenance phases for children aged 1 to younger than 10 years with lower-risk ALL.[6,71]
In a CCG study, dexamethasone was compared with prednisone during the induction and maintenance phases for children aged 1 to younger than 10 years with lower-risk ALL.[6,71]
Patients randomly assigned to receive dexamethasone had significantly fewer CNS relapses and a significantly better EFS rate.
In a MRC United Kingdom Acute Lymphoblastic Leukaemia (UKALL) trial, dexamethasone was compared with prednisolone during the induction and maintenance phases in both standard-risk and high-risk patients.[7]
Sources: 7
In a MRC United Kingdom Acute Lymphoblastic Leukaemia (UKALL) trial, dexamethasone was compared with prednisolone during the induction and maintenance phases in both standard-risk and high-risk patients.[7]
The EFS and incidence of both CNS and non-CNS relapses improved with the use of dexamethasone.
Dexamethasone was associated with an increased risk of steroid-associated toxicities, including behavioral problems, myopathy, and osteopenia.
In a DFCI ALL Consortium trial, patients were randomly assigned to receive either dexamethasone or prednisone during all postinduction treatment phases.[24]
Sources: 24
In a DFCI ALL Consortium trial, patients were randomly assigned to receive either dexamethasone or prednisone during all postinduction treatment phases.[24]
Dexamethasone was associated with a superior EFS, but also with a higher frequency of infections (primarily episodes of bacteremia) and, in patients aged 10 years or older, an increased incidence of osteonecrosis and fracture.
Triangle