A. Kamili, B. A. Hearn, I. Temelkoska, J. Solomon, M. S. Wheatley, C. H. T. Bui, C. J. Atkinson, L.-J. Elhassadi, S. Qureshi, J. Murray, R. Cadiz, A. J. Gifford, L. Cui, L. Crowley, A. Lin, C. Mayoh, M. Wong-Erasmus, T. W. Failes, G. M. Arndt, M. Haber, M. D. Norris, M. E. M. Dolman, T. N. Trahair, J. I. Fletcher
Purpose: Treatment for children with high-risk neuroblastoma relies on conventional chemotherapy and anti-GD2 immunotherapy. However, 5-year survival is only 50%, with high rates of late effects. Targeted therapy combinations are a major priority for these patients. The BCL2 inhibitor venetoclax, in combination with cyclophosphamide/topotecan, has clinical activity in relapsed and refractory neuroblastoma. We sought more effective and safer venetoclax combinations through systematic preclinical testing. Experimental design: Synergistic combinations were identified by high-throughput screening using patient-derived xenograft (PDX) models and confirmed in vivo. The leading combination (venetoclax-alisertib) was compared to combination chemotherapy in a clinical trial-like study using 22 PDX models, in scheduling experiments designed to reduce short-term toxicity, and in combination with anti-GD2 immunotherapy. BCL2 and Bim-BCL2 complex protein levels were assessed as predictors of sensitivity. Results: In vitro synergy with venetoclax was observed for standard-of-care chemotherapies and targeted agents, including DNA topoisomerase, microtubule, HDAC and Aurora kinase A (AURKA) inhibitors. Venetoclax-alisertib was particularly effective in vivo. In an n=1 study, venetoclax-alisertib induced objective response in all models. Activity was most striking in models of MYCN-amplified disease (n=12), doubling median survival time compared to cyclophosphamide/topotecan, and outperforming venetoclax-cyclophosphamide-topotecan. Efficacy was maintained with discontinuous schedules, minimizing hematological toxicity without substantially compromising activity. PDX-engrafted animals treated with venetoclax-alisertib and anti-GD2 immunotherapy survived tumor-free long-term. BCL2 expression and BCL2-Bim complex levels were of limited value for predicting response. Conclusion: Our findings support advancement of BCL2-AURKA inhibition to clinical trial for neuroblastoma with or without anti-GD2 immunotherapy, particularly in patients with MYCN amplified disease.