ABL family genes in pediatric acute lymphoblastic leukemia: Molecular determinants of tyrosine kinase inhibitor response

BCR::ABL1-positive and ABL-class acute lymphoblastic leukemia (ALL) together account for approximately 5% of pediatric and 25% of adult ALL cases. ABL-class ALL is characterized by tyrosine kinases ABL1, ABL2, PDGFRB, or CSF1R fused to various partner genes, excluding the sentinel BCR::ABL1 fusion. ABL-class ALL shares several characteristics with BCR::ABL1-positive ALL, including a similar gene expression profile, poor response to chemotherapy only, and sensitivity to tyrosine kinase inhibitors. Some patients develop tyrosine kinase inhibitor resistance through ABL1 kinase domain mutations, leading to relapse or refractory disease, but not all relapsed or refractory cases are associated with these mutations. This suggests the existence of other intrinsic tyrosine kinase inhibitor resistance mechanisms that are yet to be elucidated. For ABL-class ALL, tyrosine kinase inhibitors have been introduced more recently hence there is limited knowledge on determinants of sensitivity to tyrosine kinase inhibitors. We aimed to enhance our understanding of the determinants of tyrosine kinase inhibitors sensitivity in BCR::ABL1-positive and ABL-class ALL.

Ex vivo analysis of 32 pediatric and 19 adult BCR::ABL1-positive ALL samples from diagnosis revealed variable responses to imatinib, which correlated with responses to next-generation tyrosine kinase inhibitors, dasatinib and bosutinib. Notably, diagnostic samples from patients who relapsed after imatinib treatment exhibited greater ex vivo imatinib-resistance. Reduced BCR::ABL1 protein expression and phosphorylation were associated with intrinsic tyrosine kinase inhibitor resistance, indicating decreased dependence on BCR::ABL1 signaling. Imatinib resistance was also linked to deletions or mutations in B-cell development genes IKZF1 and PAX5. Similarly, ex vivo analysis of 16 diagnostic or relapsed pediatric ABL-class ALL samples showed varying sensitivity to imatinib, dasatinib, and bosutinib. The specific tyrosine kinase gene involved in ABL-class ALL determined tyrosine kinase inhibitor sensitivity. ABL1-fused ALL samples exhibited similar sensitivity to all three tyrosine kinase inhibitors, while PDGFRB-fused ALL samples displayed reduced sensitivity to dasatinib and bosutinib. Unlike in BCR::ABL1-positive ALL, no association was found between imatinib sensitivity and IKZF1 or PAX5 deletions, although these deletions were common in ABL-class ALL.

Besides evaluating tyrosine kinase inhibitor sensitivity, our study aimed to optimize measurable residual disease monitoring. We compared monitoring using the genomic breakpoint of the ABL-class fusion gene with the conventional method based on the clonality of leukemia-specific rearrangements of target genes immunoglobulin and T-cell receptor. The conventional method faces limitations, as approximately 5% of patients lack suitable targets, and ongoing rearrangements or subclonal presence of the target can render the method unreliable. We found an 89% correlation between the two methods. The fusion gene method measured MRD in three patients lacking conventional targets and monitored the full leukemic clone in one patient for whom part of the leukemic cells lacked the conventional targets that were selected at diagnosis. Lastly, using the oncogenic fusion gene, we discovered a patient with multi-lineage presentation rather than ALL with implications for diagnosis and treatment.

Overall, our research contributes to a deeper understanding of the biology and treatment response of ABL-family ALL, emphasizing the importance of exploring alternative treatment options and refining monitoring strategies to improve patient outcomes.