Young adult and pediatric ALL: Current induction regimens and prognostic factors

In a review published in Expert Review of Hematology, Saveria Capria and colleagues discuss the currently available induction treatments for children and adolescents with acute lymphoblastic leukemia (ALL), along with key prognostic factors.¹

Introduction

In children, the most common form of cancer is ALL. Current treatments result in an overall survival of ~ 80%, although in some high-risk subtypes this is lower. Treatment intensity stratification for childhood ALL is based on relapse risk, and certain factors have been identified to be associated with high-risk disease: age < 1 year and > 10 years, white blood cell count > 50,000−100,000 cells/μL, and the involvement of secondary organs at baseline. While T-cell phenotype used to be a high-risk factor in the past, with current treatments, the outcome is now comparable to B-cell ALL.

The identification of cytogenetic abnormalities in leukemic blasts can greatly aid risk stratification. Hyperdiploidy and ETV6/RUNX1 represent favorable markers, whereas hypodiploidy, BCR-ABL fusion, and KMT2A rearrangements are unfavorable. The use of polymerase chain reaction (PCR)/ and or flow cytometry to assess measurable residual disease (MRD) has also helped to define prognostic risk. Based on the work of pediatric oncology cooperative groups, children with favorable-risk ALL will be treated with less intensive regimens, while those with high-risk ALL still require aggressive treatment.

Induction therapy

Induction chemotherapy aims to induce a complete response, and > 95% of children with ALL achieve this. However, persistence of leukemic blasts in the blood, bone marrow, or extra medullary sites after 4–6 weeks indicates induction failure and occurs in 2−3% of patients. This group has one of the poorest prognoses in childhood ALL. Other high-risk features associated with treatment failure include older age, high leukocyte count, and 11q23 rearrangement.

Current induction regimens commonly consist of vincristine, asparaginase, and corticosteroids, and many also add an anthracycline. Anthracycline use occurred early in the history of ALL treatment and may have been responsible for an increase in the 5-year survival rate from 30% to 70%. Their use however comes with a caveat, as cardiotoxicity is a known adverse event for this drug class. Risk stratification has led to some regimens attempting to spare low-risk patients from anthracycline treatment.

Vincristine

A core component for the treatment of pediatric ALL, vincristine has been used since the 1960s. However, it is associated with the occurrence of vincristine-induced peripheral neuropathy in > 70% of patients, which decreases the quality of life of the children treated.

Glucocorticoids

Traditionally, prednisolone has been most frequently chosen for incorporation into the induction regimen and dexamethasone has been saved for the intensification phase, as it has a 6−7-fold higher efficacy in terms of anti-inflammatory action. Dexamethasone has a longer half-life in the plasma, a lower protein-bound fraction, and a longer half-life in the cerebrospinal fluid which allows greater penetration and higher concentration levels. Certain clinical trials have reported improved outcomes with dexamethasone in induction, which may be due to decreased central nervous system (CNS) relapse. However, other trials have found little difference compared with prednisolone, and often dexamethasone is associated with increased risk of osteonecrosis or infection.

Asparaginase

As ALL cells are unable to synthesize asparagine, asparaginases are an essential part of the treatment of this disease. Different sources of asparaginase are available, with not all enzymes being created equal in terms of pharmacokinetics and their stimulation of the immune system. The main sources are Escherichia (E.) coli (producing native and pegylated (PEG-)asparaginase) and Erwinia chrysanthemi. In a study from the 1990s, native E. coli asparaginase was compared with E. chrysanthemi crisantaspase at a dose of 25,000 IU/m². The results showed that crisantaspase was associated with an inferior 5-year event free survival (EFS) of 78.4% compared with 89.3% for asparaginase.

Asparaginase can also activate the immune system, resulting in anti-asparagine antibodies being produced, and this is the main reason for treatment resistance. Antibody production is rare during the induction phase but may become more of an issue at drug re-exposure. Resistance can present as signs of hypersensitivity or asparaginase can be ‘silently inactivated’. The different types of asparaginases elicit differing hypersensitivity rates, with native E. coli asparaginase being the greatest inducer of hypersensitivity (9−75%), whereas crisantaspase and PEG-asparaginase are much lower (3−37% and 4−8%, respectively). This lower immunogenicity coupled with a longer half-life means that PEG-asparaginase is increasingly used as first choice in clinical practice. Crisantaspase doesn’t display cross-reactivity with either type of asparaginase and can therefore be used as a second-line therapy.

Measurable residual disease

The strongest prognostic indicator in childhood ALL is response to chemotherapy. Assessment of the number of blasts in the peripheral blood at Day 8 and the percentage of residual blasts in the bone marrow have been used to guide risk-stratification therapy. MRD is commonly measured using PCR or flow cytometry. Flow is quicker and cheaper, but the sensitivity is reduced compared with PCR, with only 0.01% of residual leukemic cells detected versus 0.001% for PCR. Relapse risk has been significantly predicted by MRD status. MRD was measured at Day 15 in the AIEOP-BFM ALL 2000 study, and it was found that > 90% of children achieving < 0.1% residual blasts in the bone marrow remained relapse free after 5 years. MRD can also be used to identify patients that might respond to less intensive therapy. In the St. Jude Children's Research Hospital Total Therapy Study 15, flow cytometry and/or PCR were used to assess MRD at Days 19 and 46 of remission induction therapy. MRD ≥ 1% at Day 19 in patients with T-ALL, hyperdiploid B-ALL, or National Cancer Institute (NCI) standard risk ALL marked a significantly poorer outcome, with a relapse rate of 26.6% compared with 7.6% for MRD-negative patients. Out of the patients who were MRD negative, those with hyperdiploidy and ETV6-RUNX1 had a notably low relapse risk of 1.9%, indicating that for these patients, less intensive treatment could be justified.

Central nervous system prophylaxis and treatment

CNS involvement is an important issue in the treatment of ALL. Before 1970, this component was left out of induction regimens, and consequently most patients experienced CNS relapse, despite achieving bone marrow remission. To target ALL in the CNS, different therapeutic approaches are used, such as intrathecal (IT) administration, systemic chemotherapy that can bypass the blood-brain barrier, and cranial radiation. Single agents or a triple combination of methotrexate, cytarabine and hydrocortisone (intrathecal triple therapy, ITT) can be used for IT therapy and usually three doses are administered, while five doses are recommended for children with CNS involvement at diagnosis. Cranial radiation is associated with increased long-term sequelae and therefore more recent therapeutic regimens aim to omit this treatment modality. In the St. Jude Children's Research Hospital Total Therapy Study 16, systemic and IT chemotherapy without cranial radiation were used and rates of CNS relapse were low (isolated CNS relapse, 1.5%; any CNS relapse, 1.8%).

Adolescents and young adults (AYAs)

Historically, outcomes for AYAs (age range: 15–39 years) have been poorer than those of pediatric patients with only 30–45% of this group experiencing long-term EFS. The greater number of high-risk genetic alterations such as BCR-ABL, KMT2A and IGH translocations found in leukemia cells in this older population may explain the less favorable prognosis for AYA patients.

While previously AYAs have been treated with adult ALL protocols, retrospective analysis has shown that pediatric regimens can significantly improve the 5- and 7-year OS, reaching almost 70%. The main difference in the approaches are the chemotherapy agents used and/or the dose intensity. Adult protocols traditionally favor higher doses of cytarabine compared with adapted pediatric regimens which use higher cumulative doses of vincristine, asparaginase, and steroids along with greater intensity CNS therapy. However, while purely adult induction regimens have a poorer outcome in AYAs, adult protocols that utilize pediatric backbones, such a CNS prophylaxis, MRD-driven intensification of induction, and prolonged maintenance can achieve similar results compared with the purely pediatric regimens.

Philadelphia chromosome-positive ALL

The t(9;22)(q34;q11.2) translocation is commonly known as the Philadelphia chromosome and results in the formation of the BCR-ABL fusion protein. In contrast to adult ALL, the translocation is only found in 3−5% of pediatric ALL patients and previously has had a poor prognosis, with long term EFS of ~ 30%. This changed with the advent of tyrosine kinase inhibitors (TKIs), such as imatinib, with patients achieving 80% 3-year EFS in one study.

A study from 2010 found that continuous treatment from Day 15 of induction with imatinib resulted in a complete remission rate of 97%, however only 33% of patients were MRD negative after the induction phase. The high persistence of residual BCR-ABL⁺ ALL cells despite TKI therapy may explain that subsequent allogeneic hematopoietic stem cell transplantation is required to achieve long-term cure in these patients. In addition, imatinib use in conjunction with chemotherapy can induce severe toxicity, and novel, potentially less toxic TKIs, such as dasatinib and ponatinib, are under investigation.

Philadelphia-like ALL

A new sub-group of B-ALL has been recently identified, called Philadelphia-like ALL. It is more common in young men and is characterized by poor outcomes with inadequate responses to induction therapy, high relapse rates, and lower overall survival (~ 60% at 5 years) compared with BCR-ABL-negative B-ALL.

Philadelphia-like ALL makes up ≤ 15% of pediatric ALL and 20-25% of all AYA ALLs. While the expression profile of Philadelphia-like ALL is similar to Philadelphia-positive ALL, it is genetically very heterogenous. Mutations of the cytokine receptor gene CRLF2 are found in approximately half of patients along with activating-mutations of Janus activating kinase genes JAK1 and JAK2. These changes are significant with respect to potential therapeutic options; TKIs and JAK inhibitors like ruxolitinib have shown promise in this setting.

Down syndrome ALL

The risk of developing ALL is 20-fold higher in children with Down syndrome (DS), and almost 3% of ALL cases overall involve DS patients. In terms of immunophenotype, these cases are almost all B-cell precursor. DS-ALL is challenging to treat, as the etiology of the increased risk is unknown and patients have a unique toxicity profile, often associated with a higher relapse rate. High mortality rates during different phases of chemotherapy have been recorded, even during maintenance which rarely occurs in other ALL patient groups. Better supportive care and lower doses of chemotherapy are needed to improve outcomes in these patients. Targeted therapies such as blinatumomab, a CD19/CD3 bispecific antibody, and anti-CD19 chimeric antigen receptor (CAR) T-cell therapy CTL019 (tisagenlecleucel) have achieved positive results in a small population of DS-ALL patients, with high complete remission rates and a manageable safety profile.

Infant ALL

Infant ALL, characterized as disease occurring in children under 1 year of age, shows a lower incidence compared to older pediatric cases and is more aggressive. Balanced chromosomal translocations involving the KMT2A gene occurs in 70−80% of cases, while other characteristic features include a very immature B-cell phenotype, co-expression of myeloid markers, and a high tumor burden at diagnosis. Together these features negatively affect prognosis for infant ALL.

Different treatment regimens are used in infant ALL compared with pediatric patients, with an increased dependence on cytarabine in the early treatment phases as there is evidence of sensitivity of lymphoblasts to this drug. In the Interfant-99 trial, dexamethasone, vincristine, daunorubicin, native E. coli asparaginase with low-dose cytarabine, preceded by a prednisone pre-phase for 7 days, resulted in a 5-year OS of 55.2% and 5-year EFS of 46.1%. This shows that work still needs to be done in this group to improve outcomes.

Conclusion

Currently, the main challenge in pediatric ALL is the identification of low-risk patients for whom low intensity regimens might be effective, and therefore allowing the reduction of treatment-related toxicity and induction mortality rate. On the other hand, recognizing patients with unfavorable prognostic markers that will benefit from high intensity treatment is also necessary.

While the treatment of ALL in children frequently results in high complete response rates, there are still subgroups of children, such as those with DS, AYAs, patients with Philadelphia-like ALL, and infants, that have inferior outcomes. Identification of markers for high-risk disease will allow stratification of patients into intensive and less-intensive treatment groups, which will help to improve outcome. Further clinical trials with new induction treatments such as CAR T cells, targeted agents, and bispecific antibodies are showing promise but will still need to be refined to achieve comparable response rates.