Relationship between efficacy of phase I/II drug ispinesib in medulloblastoma, molecular subtypes, and p53-mutant status.

e14001 Background: KIF11, a mitotic kinesin, is responsible for assembly and maintenance of mitotic spindle during mitosis. Tumor cells can upregulate KIF11 to enhance tumor proliferation. Many cancers overexpress KIF11, including medulloblastoma (MB), the most common malignant brain tumor in children. MB is molecularly classified into 4 molecular subtypes: sonic-hedgehog (SHH)-activated MB, Wingless (WNT)-activated, and less characterized Group 3 and 4 subtypes. Group 3 and 4 MBs are the most aggressive subtypes and Group 3 MBs have the worst survival outcome. p53 mutation defines a high-risk group in MB patients. We evaluated the efficacy of a Phase I/II KIF11 inhibitor, ispinesib on MB. Methods: We employed in-vitro proliferation assays (CCK), flow cytometry (Annexin V, Propidium Iodide), western blot, Hoechst staining, and two patient-derived orthotopic xenograft (PDOX) mouse models for in-vivo drug testing. Results: Ispinesib effectively inhibited proliferation at nanomolar concentrations for all 5 MB cell lines. A comparison of IC 50 (day 4) indicated that p53-mutational-status did not influence sensitivity to ispinesib. In all five cell lines, regardless of p53 mutational status, ispinesib led to G2/M arrest. However, different p53-mutational-status led to different modes of cell death. High levels of necrotic cell death were observed in p53-mutant, while apoptosis was the predominant mode of cell death in p53-wildtype, on flow cytometry. Apoptotic nuclei were more significantly observed in p53-wildtype (p = 0.002) while p53-mutant cell line demonstrated more micronuclei (p = 0.0302) and multi-nucleation (p= 0.0161), compared to respective DMSO-control cells on Hoechst staining. Ispinesib has demonstrated effective blood brain barrier penetration in brain xenografts. We compared 2 PDOX models of MB. Ispinesib significantly improved survival outcome of Group 3 MB (Control=6, Treated=7, Log-rank p=0.0003), but worsened the survival outcome of p53-mutant SHH MB (Control=14, Treated=11, Log-rank p=0.169). This differential efficacy of ispinesib was recapitulated in the matched 3D-neurosphere cultures of both PDOX models. Cancer-stem-cell (CSC)-enriched neurospheres derived from aggressive Group 3 subtype demonstrated susceptibility to 7-day in-vitro treatment with ispinesib (AUC=0.7743). Monolayer tumor cells demonstrated similar susceptibility (AUC=0.7757). Area under curve (AUC)>0.7 was the significant cut-off. In contrast, both CSC-enriched neurospheres (AUC=0.379) and monolayer tumor cells (AUC=0.491) derived from p53-mutant SHH MB did not demonstrate susceptibility to ispinesib, consistent with matching PDOX models. Conclusions: Preclinically, our results indicate feasibility of ispinesib treatment in MB, specifically aggressive Group 3 MB. Importantly, p53-mutant SHH MB represents a poor molecular-subtype candidate for ispinesib therapy.