Introduction:
High-risk neuroblastoma (HR-NB) has one of the lowest survival rates of all paediatric cancers accounting for 15% of all paediatric cancer deaths with 60% of patients relapsing. BCL2 inhibitor venetoclax, in combination with cyclophosphamide and topotecan, has shown promising activity in an early phase trial in HR-NB patients and is a recurrent recommendation from the Australian child cancer personalised medicine trial, Zero Childhood Cancer (NCT03336931). However, this combination is limited by myelosuppression with continuous dosing. We aim to identify more effective and better tolerable venetoclax combinations using unbiased high-throughput screening and in vivo testing using patient-derived xenograft (PDX) models.
Methods:
Co-immunoprecipitation was used to identify PDX models with low, intermediate, and high levels of Bim-BCL2 protein complex, a putative biomarker of venetoclax sensitivity. Combination drug screening was conducted ex vivo in three freshly dissociated PDX models with varying level of Bim-BCL2 complex using the IC30 concentration of venetoclax with a library of >120 approved oncology drugs (0.5nM–5µM) enriched for agents with paediatric safety data. Synergistic combinations were tested for tolerability in NSG mice, and then for efficacy in the PDX models in vivo, using schedules approximating those achievable in patients.
Results:
Drugs synergistic with venetoclax in combination screens included conventional chemotherapies (DNA topoisomerase and microtubule inhibitors) and targeted agents (PARP, HDAC and AURKA inhibitors). Three combinations have so far been assessed in xenograft models. Vincristine-venetoclax prolonged survival in all models tested, while vorinostat-venetoclax had limited efficacy. The most effective combination, alisertib-venetoclax, elicited complete responses in all models, and outperformed venetoclax-cyclophosphamide-topotecan in one model in which studies have been completed. Furthermore, Bim-BCL2 protein complex levels predicted in vivo response to venetoclax. Additional combinations are undergoing testing. We will further assess promising combinations using less myelosuppressive discontinuous dosing schedules that are more clinically relevant and will test efficacy across of panel of HR-NB PDX models using clinical trial-like (n=1) experimental designs.
Conclusion:
Systematic preclinical testing can be utilised effectively to identify the most synergistic venetoclax combination. Alisertib-venetoclax is a highly promising combination based on preclinical studies, and should be considered for early-phase clinical trials.