Impact of asparaginase discontinuation on outcomes in pediatric B-cell precursor acute lymphoblastic leukemia: Post hoc analysis of the JCCG ALL-B12 Study

Abstract Background Asparaginase (Asp) is a key component in the treatment of children with acute lymphoblastic leukemia (ALL), particularly in the induction and consolidation phases. However, hypersensitivity and other toxicities frequently lead to premature discontinuation, thus raising concerns about its impact on long-term outcomes. A nationwide clinical trial, the ALL-B12 study, was conducted in Japan between 2012 and 2017 to investigate the effectiveness of unvalidated treatment phases in each risk group of pediatric patients aged 1–19 with B-cell precursor ALL (BCP-ALL), through randomized controlled trials and establish a safety-focused treatment framework for BCP-ALL. The main results of ALL-B12 were reported in 2024. The current study aimed to evaluate the clinical implications of L-Asp truncation in this large cohort of children with BCP-ALL treated under the ALL-B12 protocol in Japan. Methods A total of 1,804 patients were included in the full analysis set of the ALL-B12 study. Patients with NCI-SR and NCI-HR were assigned to the standard risk group (SR) and the intermediate risk group (IR), respectively. Poor steroid responders and those with high-risk genomic abnormalities were assigned to the high risk group (HR). In the current analysis, patients were categorized based on whether they completed the full L-Asp schedule (complete group) or discontinued prematurely (discontinuation group). A subset of HR patients who discontinued L-Asp received vincristine and dexamethasone pulses to compensate for the treatment reduction. For SR/IR patients no additional modification was applied when L-Asp was discontinued. Clinical characteristics and outcomes were compared between these two groups. We conducted a landmark analysis of patients who remained event-free at key treatment landmarks in each group—the start of maintenance therapy for all patients, and the date of stem cell transplantation only for HR patients who received transplantation. Multivariable analysis was performed using a Cox regression model with follow-up starting from the end of induction, treating L-Asp discontinuation as a time-dependent covariate. In Japan, Erwinia asparaginase was not approved for use and could not be used as a substitute for native L-asparaginase in case of discontinuation during the study period. Asparaginase activity levels were not measured in this study. Results Among the 1,804 patients, 142 (7.9%) discontinued L-Asp prematurely. The most frequent reason for discontinuation was acute pancreatitis (AP, n = 81, 4.5% of all patients), followed by clinical hypersensitivity (CH, n = 55, 3.0%). Patients with AP discontinued L-Asp earlier: 53 /81 experienced events in the induction phase, whereas only 5/55 patients with CH experienced events in the induction phase. The analysis of clinical characteristics suggested that the discontinuation group was older at diagnosis (median 7 years old vs. 4 years old), and was thus more likely to be classified as NCI-HR compared to the complete group. Landmark analyses revealed that the 5-year event-free survival (5yEFS) were 82.4% vs. 91.1% among SR patients (p = 0.087), 70.7% vs. 83.9% among IR patients (p = 0.054), and 75.9% vs. 82.9% among HR patients (p = 0.346) in the discontinuation and complete groups, respectively. While discontinued earlier, the EFS was not different between those with AP and CH (5yEFS, 75.9% vs. 76.3%; p = 0.882). According to the Cox regresssion analysis, the effects of L-Asp discontinuation on outcomes differed across subgroups. Those who showed a poor response to the treatment (prednisolone poor responder, day 15 bone marrow blast < 25%, or MRD < 10-4 at the end of induction or at the end of consolidation) had worse outcomes when L-Asp was discontinued. Mulitivariable analysis including clinical risk factors (age/CNS status/WBC count at diagnosis, and MRD status at the end of induction) revealed a significant association between L-Asp discontinuation and EFS (HR 1.55; 95% CI, 1.02–2.36; p = 0.041). Discussion We found that BCP-ALL patients with truncated L-Asp treatment had worse EFS than those who completed L-Asp treatment, which may be partly attributed to the older age at diagnosis in the discontinuation group. Given that the intensified L-Asp arm did not show a significant benefit in the randomized controlled study within ALL-B12, our data support the existence of an appropriate and balanced dose of asparaginase tailored to the specific needs of each risk group.