IKZF1 Deletions Induce Resistance to AraC in BCP-ALL Reversible By CDK8/19 Inhibition

Background Deletions and mutations affecting lymphoid transcription factor IKZF1 occur in about 10-15% of pediatric patients and up to 50% of adults with B cell precursor acute lymphoblastic leukemia (BCP-ALL) and predict a poor outcome. We have shown previously that loss of IKZF1 function compromises the therapeutic efficacy of glucocorticoids (GCs). Here, we investigated whether loss of IKZF1 function also affects therapy response to other chemotherapeutic agents used in the treatment of ALL.

Aim To develop tailored treatment protocols for IKZF1 deficient leukemia

Results We used CRISPR-Cas9 to target the IKZF1 locus in SEM pro-B ALL cells and evaluated the response to standard chemotherapy used in the treatment of ALL. We observed that IKZF1^-/- SEM cells show a markedly reduced response to the pyrimidine analogue cytarabine (AraC) relative to IKZF1 wild type cells, with an 8-fold increase in IC⁵⁰. The same drug resistance profile was observed upon shRNA-mediated knockdown of IKZF1 as well as in B-cells isolated from 8 to 14 weeks-old Ikzf1^+/- mice.

Using ex vivo cultures of patient derived xenografts (PDX), we observed that IKZF1 deficient leukemic cells were significantly more resistant to AraC relative to cells wild type for IKZF1 (3-fold increase in area under the curve (AUC) (Figure A). In addition, we targeted the IKZF1 protein for proteasomal degradation using the thalomide Iberdomide in PDX cells wildtype for IKZF1, showing that this drug-induced IKZF1 loss leads to AraC resistance in 7 out of 8 PDX samples.

To provide in-patient evidence for AraC resistance, we compared the reduction in tumor load between patients with leukemias wildtype for IKZF1 with those carrying IKZF1 deficiencies after treatment with an AraC containing treatment block (cumulative dose of 1200 mg/m2 AraC, 6-mercaptopurine 60 mg/m2/day p.o., for 28 days and 2 doses of 1000 mg/m2 cyclophosphamide). In the Dutch ALL10/11 protocol, this treatment block is flanked by minimal residual disease (MRD) measurements. We selected all patients with a high tumor load (MRD of 10^-2) before the start of the AraC containing block and calculated the reduction in leukemia burden by comparing MRD levels before and after treatment with the AraC containing interval. This showed that the response to AraC treatment was attenuated in patients with IKZF1 deleted ALL (n=25) compared to patients WT for IKZF1 (n=35) (**p=0.0065).

Because IKZF1 deficient cells did not show resistance to other compounds used in this treatment block, we attribute the reduced treatment response to increased AraC resistance. To further confirm the involvement of IKZF1 loss in AraC resistance, we performed a single mouse trial in which we compared AraC treatment response in mice transplanted with IKZF1 wildtype leukemias (n=5) with IKZF1 deficient leukemias (n=7), confirming that the latter respond more poorly to treatment.

To study whether resistance in response to IKZF1 loss can be reversed and obtain more insights into the biology behind the observed AraC resistance, we performed CRISPR/Cas9 screens to identify drug targets that can restore response to AraC therapy. Briefly, IKZF1 deficient cells were transduced with a library of gRNAs targeting subset of genes (kinases and genes involved in cellular metabolism). Transduced cells were treated with a sublethal dose of AraC and the relative abundance of each gRNA before and after treatment was calculated. In addition to known modifiers of the AraC therapy response such as deoxycytidine kinase, we also identified novel targets that may be targeted to sensitize cells to cytarabine treatment. Guide RNAs targeting the mediator kinases CDK8/19 were selectively and strongly depleted during treatment, suggesting that these proteins modify response to cytarabine treatment. Indeed, small-molecule inhibitors targeting CDK8/19 kinase activity synergized with cytarabine induced killing of leukemic cells (Figure B).

Conclusion Together, our results demonstrate that loss of IKZF1 function confers resistance to AraC in BCP-ALL. Moreover, this resistance can be overcome by use of small molecule inhibitors such as those targeting CDK8/19. We expect that these studies will aid in the rational design of combination therapies that will further improve treatment response in these high-risk leukemias.

van der Velden:Cytognos: Other: laboratory services agreement; Agilent: Other: laboratory services agreement; Euroflow: Patents & Royalties; BD biosciences: Other: laboratory services agreement.