Ruxolitinib (rux) Phase II clinical trials are underway for the treatment of high-riskJAK2-rearranged (JAK2r) B-cell acute lymphoblastic leukemia (B-ALL). Treatment resistance to targeted inhibitors in other settings is common; elucidating potential mechanisms of rux resistance inJAK2r B-ALL will enable development of therapeutic strategies to overcome or avert resistance. We generated a murine pro-B cell model ofATF7IP-JAK2with acquired resistance to multiple type-I JAK inhibitors. Resistance was associated with mutations within theJAK2ATP/rux binding site, including aJAK2p.G993A mutation. Using in vitro models ofJAK2r B-ALL,JAK2p.G993A conferred resistance to six type-I JAK inhibitors and the type-II JAK inhibitor, CHZ-868. Using computational modeling, we postulate thatJAK2p.G993A enabled JAK2 activation in the presence of drug binding through a unique resistance mechanism that modulates the mobility of the conserved JAK2 activation loop. This study highlights the importance of monitoring mutation emergence and may inform future drug design and the development of therapeutic strategies for this high-risk patient cohort.

High-risk Ph-like ALL includes genomic rearrangement of the ABL1 and ABL2 genes (collectively ABL-rearranged, ABLr), and novel treatments are required. For High-risk Ph-like ALL includes genomic rearrangement of the ABL1 and ABL2 genes (collectively ABL-rearranged, ABLr), and novel treatments are required. For

B-cell acute lymphoblastic leukaemia (B-ALL) is characterised by diverse genomic alterations, the most frequent being gene fusions detected via transcriptomic analysis (mRNA-seq). Due to its hypervariable nature, gene fusions involving the Immunoglobulin Heavy Chain (IGH) locus can be difficult to detect with standard gene fusion calling algorithms and significant computational resources and analysis times are required. We aimed to optimize a gene fusion calling workflow to achieve best-case sensitivity for IGH gene fusion detection. Using Nextflow, we developed a simplified workflow containing the algorithms FusionCatcher, Arriba, and STAR-Fusion. We analysed samples from 35 patients harbouring IGH fusions (IGH::CRLF2 n = 17, IGH::DUX4 n = 15, IGH::EPOR n = 3) and assessed the detection rates for each caller, before optimizing the parameters to enhance sensitivity for IGH fusions. Initial results showed that FusionCatcher and Arriba outperformed STAR-Fusion (85–89% vs. 29% of IGH fusions reported). We found that extensive filtering in STAR-Fusion hindered IGH reporting. By adjusting specific filtering steps (e.g., read support, fusion fragments per million total reads), we achieved a 94% reporting rate for IGH fusions with STAR-Fusion. This analysis highlights the importance of filtering optimization for IGH gene fusion events, offering alternative workflows for difficult-to-detect high-risk B-ALL subtypes.

• Asciminib inhibits Abl2 (Abelson-related gene) by interacting with different myristate pocket residues when compared with Abl. • ABL2 exon 3–encoded region of the SH3 domain is critical for STAMP inhibitor activity against ABL2 rearrangements in vitro and in vivo.