Acute lymphoblastic leukemia (ALL) is the most common pediatric cancer.1 Tremendous improvements in survival for children with ALL have been achieved in the past 50 years, with survival rates exceeding 90 percent.1
AALL0331 was a Children’s Oncology Group (COG) cooperative trial that enrolled National Cancer Institute standard risk (SR; see Table 1 for definitions) patients between 2005 and 2010.2 All patients received a three-drug induction with dexamethasone, vincristine, and pegaspargase-asparaginase (PEG-ASP). Following induction, patients were risk stratified into three categories: SR-low, SR-average, or SR-high. This article describes the outcome for those with SR-average and SR-high disease.2
Risk stratification was based on central nervous system status, cytogenetics, and response to therapy (Table 1). Patients were evaluated for response to therapy via bone marrow aspirate on day 8. If blasts were over 5 percent, the patients had a repeat bone marrow aspirate on day 15. Bone marrow response by morphology and minimal residual disease (MRD) were also assessed after induction (day 29). Patients were classified as rapid early responders or slow early responders based on marrow response (Table 1). Patients with MRD greater than 1 percent or an M2 marrow at day 29 received two more weeks of therapy and were also classified as slow early responders if they had MRD less than 1 percent at day 43. Those with M3 marrow at day 29 or who had not achieved a remission (M1) by day 43 were removed from protocol therapy.
Definitions and Risk Stratification Used in AALL0331
| NCI Standard Risk (SR) | Age between 1 and 9.99 years and initial white blood cell count < 50,000/mL |
| Rapid Early Response (RER) | Bone marrow blasts <5% by morphology by day 15 and day 29 minimal residual disease (MRD) <0.1% |
| Slow Early Response (SER) | Bone marrow blasts ≥5% on day 15 and/or day 29 MRD ≥0.1 and <1%. |
| M1 | <5% bone marrow blasts by morphology |
| M2 | 5-25% bone marrow blasts by morphology |
| M3 | >25% bone marrow blasts by morphology |
| Central nervous system (CNS) 1 | No cerebrospinal fluid. (CSF) blasts on cytospin |
| CNS2 | CSF WBC <5/mL with blasts* |
| CNS3 | CSF WBC ≥5/mL with blasts on cytospin and/or clinical signs of CNS involvement* |
| Favorable Cytogenetics | Triple trisomy of chromosome 4, 10, and 17 or ETV6-RUNX1 |
| Unfavorable Genetics | BCR-ABL1 or KMT2A-R or hypodiploidy with chromosomes <44 |
| SR-Low | RER and CNS1 and favorable cytogenetics |
| SR-Average | RER and CNS1 or CNS2 and no unfavorable genetics |
| SR-High | CNS3 or KMT2A-R or SER or steroid pretreatment |
| *Traumatic lumber punctures were classified as CNS2 or CNS3 based on Steinherz/Bleyer algorithm3 | |
| NCI Standard Risk (SR) | Age between 1 and 9.99 years and initial white blood cell count < 50,000/mL |
| Rapid Early Response (RER) | Bone marrow blasts <5% by morphology by day 15 and day 29 minimal residual disease (MRD) <0.1% |
| Slow Early Response (SER) | Bone marrow blasts ≥5% on day 15 and/or day 29 MRD ≥0.1 and <1%. |
| M1 | <5% bone marrow blasts by morphology |
| M2 | 5-25% bone marrow blasts by morphology |
| M3 | >25% bone marrow blasts by morphology |
| Central nervous system (CNS) 1 | No cerebrospinal fluid. (CSF) blasts on cytospin |
| CNS2 | CSF WBC <5/mL with blasts* |
| CNS3 | CSF WBC ≥5/mL with blasts on cytospin and/or clinical signs of CNS involvement* |
| Favorable Cytogenetics | Triple trisomy of chromosome 4, 10, and 17 or ETV6-RUNX1 |
| Unfavorable Genetics | BCR-ABL1 or KMT2A-R or hypodiploidy with chromosomes <44 |
| SR-Low | RER and CNS1 and favorable cytogenetics |
| SR-Average | RER and CNS1 or CNS2 and no unfavorable genetics |
| SR-High | CNS3 or KMT2A-R or SER or steroid pretreatment |
| *Traumatic lumber punctures were classified as CNS2 or CNS3 based on Steinherz/Bleyer algorithm3 | |
Patients with SR-average disease were randomly assigned in a 2×2 factorial design to one of four treatment conditions (Table 2). In addition to the standard consolidation (SC) of vincristine, mercaptopurine, and intrathecal methotrexate, intensified consolidation (IC) included 1 g/m2 of cyclophosphamide, an increased number of doses of vincristine, cytarabine, two doses of PEG-ASP, and a lower-dose interrupted schedule of mercaptopurine. Initially, patients in the standard interim maintenance arms received oral methotrexate. Augmented interim maintenance included Capizzi methotrexate and two additional doses of PEG-ASP. Augmented delayed intensification included more doses of vincristine and a second dose of PEG-ASP. Following an amendment in 2008 based on the results of CCG-1991, all patients in the SR-average arm received escalating Capizzi methotrexate and the standard delayed intensification, reducing the randomization to between SC and IC only. Patients with SR-high disease were nonrandomly assigned to receive intensified consolidation and two cycles of delayed intensification.
Original 2 × 2 Factorial Design of Treatment Conditions for Patients With Standard-Risk Average Disease
| SSStandard consolidation Standard interim maintenance Standard delayed intensification | SCStandard consolidation Augmented interim maintenance Augmented delayed intensification |
| ISIntensified consolidation Standard interim maintenance Standard delayed intensification | IAIntensified consolidation Augmented interim maintenance Augmented delayed intensification |
| SSStandard consolidation Standard interim maintenance Standard delayed intensification | SCStandard consolidation Augmented interim maintenance Augmented delayed intensification |
| ISIntensified consolidation Standard interim maintenance Standard delayed intensification | IAIntensified consolidation Augmented interim maintenance Augmented delayed intensification |
The study enrolled 5,377 patients and included a two-stage consent. Patients initially consented to induction therapy. A second consent was used for postinduction therapy and randomization. A total of 3,992 patients continued on to postinduction therapy; 86 percent of the 1,052 patients who ended protocol-directed therapy at the end of induction did so because they or their caregivers refused further protocol therapy. The 6-year event-free survival and overall survival (OS) for all patients were 88.96 percent and 95.54 percent, respectively.
In the SR-average arm, 1,500 patients were randomized between SC and IC. There was no difference in continuous complete remission rate (CCR) or OS between these groups. More recently, COG has redefined SR-average as day 8 peripheral blood MRD less than 1 percent and day 29 MRD less than 0.01 percent. The authors analyzed the difference between the SC and IC arms in the subgroup of patients who would meet this more stringent definition of SR-average and found no benefit of IC over SC. In those patients with day 29 MRD between 0.01 and 0.1 percent, patients had a worse event-free survival and OS overall, and outcomes were equivalent between IC and SC. In all subgroups, IC was associated with significantly higher toxicity, especially infectious toxicities, with an infectious rate of 23 percent in the IC groups and 4.7 percent in the SC group (p<0.0001).
The study enrolled 635 patients on the SR-high arm, and these patients also had excellent outcomes, with OS greater than 90 percent. CCR and OS rates in patients on the SR-high arm who were treated nonrandomly with intensified therapy were superior to those patients with low-level MRD who were treated with IC or SC and standard postconsolidation therapy. The 6-year CCR rate in the SR-average low-level MRD group overall was 77.46 percent, versus 85.55 percent in the SR-high group.
In Brief
Historically, intensification of therapy has led to improved outcomes in many patients with ALL. The results of the SR-average arm on AALL0331 demonstrate that for many patients with ALL, further intensification of therapy adds only toxicity without benefit. In contrast, the SR-high patients seemed to benefit from nonrandom intensification of therapy post-consolidation. Thus, there may be some groups of patients that would continue to benefit from further intensification of therapy. It is becoming increasingly difficult to identify who those patients are. The current generation of COG trials (NCT03914625, NCT03959085, NCT03876769) have changed the paradigm and are testing whether the incorporation of immunotherapies into cytotoxic chemotherapy backbones will improve survival without additional toxicity. Time will tell if this approach will lead to superior outcomes.
References
Competing Interests
Dr. Diorio and Dr. Teachey indicated no relevant conflicts of interest.
