Children are receiving only a fraction of the advances in new adult cancer treatments. Whilst cancer survival rates for children have improved significantly over the last 20 years, for some specific cancers we have not seen a substantial improvement in cure rates over the same time period. Over the past 30 years only 6 new drugs have been approved specifically for use in paediatric cancer. There is a clear lack of incentives to support the development of specific childhood cancer drugs. In this context, drug repurposing strategies are well suited to deliver tangible benefits to paediatric cancer patients. Several examples of drug repurposing in the field of paediatric oncology/haematology, using different scientific strategies, are highlighted. More recently artificial intelligence/machine learning-powered approaches have been earning ground and one such case is presented here more in detail.
Neuroblastoma (NB) is a type of cancer that develops from immature nerve cells and subset of neuroblastoma patients has a relatively high rate of relapse and becomes refractory to treatment. There is an urgent need for new therapies to address this medical need. We aimed at the identification and validation or approved drugs that may be able to treat high risk NB in children, accelerating children’s access to new drugs and reduce the cost of the development process.
An Artificial Intelligence (AI) approach based on Disease-Gene Expression Matching (DGEM) was used to identify drug repurposing candidates. In silico analysis identified 12 already approved drugs that were further tested in vitro using patient-derived cell lines and organoids, and in vivo with NB PDX mice. Four of these compounds exhibited anti-proliferative activity and were selected for further characterization. We screened combinations of two classes of drugs and observed high synergistic effects and a substantial decrease in neuroblastoma cell viability. Combination treatment resulted in the upregulation of pathways linked to neuron projection development, metabolism of lipids, vesicle-mediated transport, and autophagy regulation. Additionally, the combination resulted in a phenotypic switch to a more chemosensitive cell state. In vitro pretreatment with the combinations confirmed the sensitization of patient-derived NB organoids to cisplatin. Preliminary In vivo experiments further support these findings.
This projects represent an example of how AI-driven approaches, combined with advanced in vitro/in vivo testing capabilities may lead to the rapid and cost-effective identification of drug repurposing candidates for rare diseases that are not prioritized by the pharma industry.
Smart Citations