Identifying Drug-Resistant Mutations in Ebf1-Pdgfrb Ph-like Acute Lymphoblastic Leukemia

Background: Advances in cancer genomics have recently identified a particular group of patients who display a gene expression profile (GEP) similar to that of Philadelphia-chromosome positive (Ph+) acute lymphoblastic leukemia (ALL) patients in approximately 15% of children and over 25% of young adults with B-ALL; thus known as Ph-like ALL. The latter has a worse prognosis compared to those without the Ph-like GEP with conventional chemotherapy. Recent studies have unraveled the genomic landscape of Ph-like ALL which is characterized by genetic alterations activating kinase signaling pathways predicted to respond to tyrosine kinase inhibitors (TKIs). In light of the remarkable outcomes of Ph+ALL patients through incorporation of TKI, the Children9s Oncology Group (COG) ALL Committee is actively working to incorporate dasatinib for Ph-like ALL patients harboring ABL-class kinase fusions (ABL1, ABL2, PDGFRB, CSF1R) and eventually, ruxolitinib for those with lesions that are predicted to respond to JAK inhibition. While it is hoped that many of these patients will be cured with the addition of relevant TKIs to chemotherapy, we hypothesize that a proportion of patients will develop resistance to TKI, similar to adults with chronic myeloid leukemia who have been treated with long-term TKI. Hence, investigating the underlying mechanisms governing therapy resistance in Ph-like ALL becomes critical for proactively developing novel therapeutic strategies in the relapsed setting. Objectives: To identify the full spectrum of mutations conferring resistance to clinically-active TKIs in Ph-like ALL and to characterize their relative biochemical resistance to different TKIs. Methods: We first focused on the EBF1-PDGFRB rearrangement since this is the most common recurrent kinase-activating fusion genes in pediatric Ph-like ALL. We used a previously validated in-vitro saturation mutagenesis screen to predict the full spectrum of EBF1-PDGFRB drug-resistant mutations. In brief, EBF1-PDGFRB plasmid was propagated into DNA-repair-deficient E. Coli strain XL-1 Red to generate a library of random mutants. Mutagenized EBF1-PDGFRB plasmid was transfected into 293T cells. Viral supernatants were collected and used to infect Ba/F3 cells. Transduced Ba/F3 cells were plated in 1.2% Bacto-agar and exposed to different TKIs (imatinib, dasatinib) at various concentrations. Genomic DNA from IL-3 independent and TKI-resistant colonies was isolated. The PDGFRB kinase domain was amplified and bidirectional sequencing was performed. Results: Our in-vitro screens showed that the vast majority of drug-resistant clones harbor a kinase domain (KD) point mutation. The predominant KD point mutation conferring resistance to imatinib (94%; 168/178 colonies) or dasatinib (81%; 338/416 colonies) was T681I, which is analogous to BCR-ABL1 T315I gatekeeper mutation. N666S was the second most common KD mutation (6%; 18/321 colonies). The full panel of KD mutations recovered is shown in Table 1. Ba/F3 cells harboring mutant EBF1-PDGFRB T681I was 100 times more resistant to dasatinib compared to wild-type and could be rescued by ponatinib, as predicted (Figure 1). Conclusion: Our screens suggest that KD point mutations may represent the primary mechanism of acquired TKI resistance in EBF1-PDGFRB Ph-like ALL. T681I was the most common KD point mutation in EBF1-PDGFRB upon exposure to imatinib or dasatinib. Future efforts should focus on targeting the T681I gatekeeper mutation with ponatinib or alternative agents for relapsed Ph-like ALL patients harboring these mutations. Disclosures Smith: Astellas: Research Funding; Plexxikon: Research Funding. Mullighan: Cancer Science Institute: Membership on an entity9s Board of Directors or advisory committees; Amgen: Honoraria, Speakers Bureau; Incyte: Consultancy, Honoraria; Loxo Oncology: Research Funding. Shah: Bristol-Myers Squibb: Research Funding; Pfizer: Research Funding; Plexxikon Inc.: Research Funding.