Are we ABL to do better for children with BCR–ABL1-like acute lymphocytic leukaemia?

Just one decade ago, two paediatric leukaemia-focused research teams independently identified what is now known as BCR–ABL1-like (also known as Philadelphia chromosome-like) acute lymphocytic leukaemia among children with high-risk, cytogenetically normal B-cell acute lymphocytic leukaemia, who showed a very high risk of relapse with conventional chemotherapy.^1,2 The leukaemic cells of these patients were found to harbour a kinase-activated gene expression profile similar to that of BCR–ABL1 fusion-positive acute lymphocytic leukaemia, yet they were negative for the sentinel BCR–ABL1 rearrangement. Over the past 10 years, detailed molecular characterisation of BCR–ABL1-like acute lymphocytic leukaemia via national and international collaborative efforts has revealed the remarkable genetic heterogeneity and characteristic constitutive kinase signalling pathways of this disease type³ and has elucidated its incidence across the paediatric-to-adult age spectrum. Numerous preclinical studies have further indicated that BCR–ABL1-like acute lymphocytic leukaemia cells show sensitivity to targeted tyrosine-kinase inhibitors in vitro and in vivo, which has led to current early-phase clinical trials assessing the safety and potential efficacy of tyrosine-kinase inhibitors in addition to chemotherapy in children and adults with relapsed or newly-diagnosed BCR–ABL1-like acute lymphocytic leukaemia.⁴ However, well defined clinical outcomes of retrospectively-identified children with BCR-ABL1-like acute lymphocytic leukaemia, who have been treated uniformly with the best available chemotherapy regimens, remain poorly defined. This absence of data has created appreciable challenges for potential efficacy assessment in current single experimental treatment group clinical trial designs, which have been required to date given the relative rarity of BCR–ABL1-like acute lymphocytic leukaemia, and comparison with appropriate historic controls.

In The Lancet Haematology, Monique den Boer and colleagues⁵ from the Ponte di Legno group, which represents more than 20 paediatric acute lymphocytic leukaemia consortia, combined forces to carefully delineate the clinical characteristics and outcomes of all identified children with ABL1-class fusion B-cell acute lymphocytic leukaemia enrolled in clinical trials done between 2000 and 2018, in which patients had been treated with multiagent chemotherapy as a first-line treatment without tyrosine-kinase inhibitors. Even though the BCR–ABL1-like subtype comprises approximately 5–8% of children with National Cancer Institute (NCI)-defined standard risk,⁶ approximately 10–15% of children with NCI-defined high risk,⁷ and up to 30% of young adults with B-cell acute lymphocytic leukaemia,⁸ ABL-class fusion rearrangements involving ABL1, ABL2, CSF1R, and PDGFRB are relatively uncommon and collectively account for approximately 10% of all BCR–ABL1-like acute lymphocytic leukaemia cases. den Boer and colleagues⁵ unsurprisingly confirm a high cumulative incidence of relapse and poor event-free survival in children with ABL-class fusion B-cell acute lymphocytic leukaemia, but also provide important new rearrangement-specific metrics (64 [52%] of 122 patients were PDGFRB fusion-positive, 40 [33%] were ABL1 fusion-positive, ten [8%] were CSF1R fusion-positive, and eight [7%] were ABL2 fusion-positive) that would have been impossible to elucidate without international collaboration. Importantly, the investigators report that levels of minimal residual disease at the end of induction therapy were high (defined as ≥10⁻² cells) in 61 (66%) of 93 patients with ABL-class fusion B-cell acute lymphocytic leukaemia, and particularly in those with ABL2 fusion-positive (six [86%] of seven patients) and PDGFRB fusion-positive rearrangements (43 [88%] of 49 patients). Minimal residual disease at the end of induction therapy of 10⁻² cells or more was expectedly predictive of an unfavourable outcome (hazard ratio of event-free survival in patients with a minimal residual disease of ≥10⁻² cells vs those with a minimal residual disease of <10⁻² cells 3·33 [95% CI 1·46–7·56], p=0·0039). Even though a previous study on BCR–ABL1- like acute lymphocytic leukaemia showed equivalently poor survival of paediatric patients with the most common CRLF2 rearrangements when compared with those who had non-CRLF2 kinase fusions (eg, ABL-class fusions, JAK2 fusions, and EPOR fusions),⁹ this study⁵ convincingly showed differential outcomes among patients treated with conventional chemotherapy regimens in the ABL-class cohort; 5-year event-free survival was 52·9% (95% CI 41·5–67·5) in those who were PDGFRB fusion-positive and 37·5% (15·3–91·7) in those who were ABL2 fusion-positive, compared with 68·6% (54·5–86·3) in those who were ABL1 fusion-positive and 80·0% (58·7–100·0) in those who were CSF1R fusion-positive. Variable initial prednisone prophase responses were observed among the four groups of patients with different ABL-class fusion types, in whom data were available; a poor response to prednisone appeared to align with a higher incidence of minimal residual disease at the end of induction therapy in patients with ABL2 fusion-positive and PDGFRB fusion-positive B-cell acute lymphocytic leukaemia compared with those who were ABL1 fusion-positive and CSF1R fusion-positive, but a poor response to prednisone was not observed to be an independent predictor of an unfavourable outcome. Co-occurring IKZF1 and other transcription factor gene deletions were frequently detected as anticipated; although, these alterations also did not appear to convey additional prognostic significance among studied patients. Finally, even though over a third (43 [37%] of 115) of the study population underwent allogeneic haematopoietic stem cell transplantation during first remission (usually because of induction failure¹⁰ or persistent minimal residual disease at the end of consolidation therapy), with an observed reduction in the 5-year cumulative incidence of relapse, 5-year event-free survival and 5-year overall survival were not improved with this approach, probably due to the high transplantation-related mortality.

This important study⁵ highlights the collaborative spirits of paediatric haematology-oncology in the study of uncommon and newly-defined genetic subtypes of childhood acute lymphocytic leukaemia, with overarching goals of improving our understanding of underlying biological pathways and optimising therapeutic approaches. This study⁵ establishes a new and necessary baseline outcomes reference in children with ABL-class fusion B-cell acute lymphocytic leukaemia treated with chemotherapy, and highlights the continued urgent need for alternative therapies that can overcome chemoresistance, decrease the risk of relapse, and increase long-term survival. Current and planned clinical trials of tyrosine-kinase inhibitors in children with BCR–ABL1-like acute lymphocytic leukaemia will undoubtedly benefit from the robust comparator outcome metrics established by this study,⁵ as we strive together to cure all children with acute lymphocytic leukaemia.