Re-Induction Outcome for Pediatric Patients with Relapsed or Refractory B-Cell Precursor Acute Lymphoblastic Leukemia: A Retrospective Cohort Study of the Therapeutic Advances in Childhood Leukemia Consortium

Introduction

Remission induction rates after a second or greater relapse is a critical endpoint in phase II trials of childhood acute lymphoblastic leukemia (ALL). A robust benchmark is crucial for identification of novel multi-agent regimens worthy of further study. The Therapeutic Advances in Childhood Leukemia and Lymphoma (TACL) consortium previously reported the response rates of children with multiply relapsed and refractory (R/R) ALL treated between 1995 and 2004, which provided a benchmark for clinical trials. To define more recent treatment patterns and test the robustness of this benchmark, we performed a retrospective cohort review of children with R/R ALL who experienced second or greater treatment failure at TACL consortium sites between 2005 and 2013.

Patients and Methods

Eligible patients were identified at participating TACL institutions. This cohort was comprised of patients with medullary R/R B-cell precursor ALL who experienced at least 2 treatment failures or relapsed after hematopoietic stem cell transplant. Patient demographic data and details of the initial and R/R disease characteristics were abstracted from medical records and entered into a central database. This study was approved by the IRB of each participating institution.

Treatment failure was defined by the presence or re-emergence of circulating blasts, M2/M3 BM, or extramedullary (EM) disease despite therapy. Complete remission (CR) was defined as M1 marrow, no EM disease and evidence of peripheral count recovery. For the purpose of statistical analysis, patients who met these criteria without platelet recovery (CRp) or normal blood count recovery (CRi) were included as CR. Univariate and multivariate logistic regression were utilized to evaluate the risk of re-induction failure. Predictors included in this preliminary analysis were NCI risk criteria at diagnosis, duration of the prior remission, the treatment attempt number, and the EM and BM status at the start of each therapy attempt.

Results

This report includes 214 patients. Fifty-six percent were male. At initial diagnosis, 32% were at least 10 years old, 26% had initial white blood cell (WBC) counts over 50,000/µL, and 39% were classified as high risk by the NCI risk criteria (Table 1). Therapy involved various combinations of agents and ranged between 2 and 10 attempts. The CR rate was 42% for third treatment attempt and 24% for fourth and subsequent treatment attempts (Table 2). Treatment failures were significantly associated with increased number of treatment attempts (p < 0.001), shorter duration of previous CR (p < 0.001) and NCI risk category at diagnosis (p = 0.018).

Conclusion

This preliminary analysis found similar CR rates in patients with third treatment failure compared to the 1^st TACL retrospective study of the prior decade (42% vs. 44%, Ko et al, 2010) and an Austrian report with a small cohort of patients (Reismüller et al, 2013). Further analysis will be performed in comparison to the initial TACL retrospective study cohort once enrollment to this study has been completed (approximately 400 patients). A robust, contemporary historical control may serve as an alternative to a randomized control when outcome with past therapies in unacceptably poor.