Poster Presentation Lowy Cancer Symposium 2015

Modelling personalised medicine in high-risk neuroblastoma (#125)

Alvin Kamili 1 , Mitchell J Lockwood 1 , Carol Wadham 1 , Tim W Failes 1 2 , Greg M Arndt 1 2 , C Patrick Reynolds 3 , Murray D Norris 1 , Michelle Haber 1 , Toby N Trahair 1 4 , Jamie I Fletcher 1
  1. Children's Cancer Institute, Lowy Cancer Research Centre, Randwick, NSW, Australia
  2. ACRF Drug Discovery Centre for Childhood Cancer, Lowy Cancer Research Centre, Randwick, NSW, Australia
  3. Texas Tech University Health Sciences Centre, Lubbock, Texas, USA
  4. Kids Cancer Centre, Sydney Children's Hospital, Randwick, NSW, Australia

Background: Neuroblastoma is the most common extracranial solid tumour of children and accounts for 15% of all paediatric oncology deaths. The cure rates of children with high-risk neuroblastoma remain less than 50%, with dose limiting toxicities and the adverse effects of chemotherapy presenting major challenges. These challenges may be partially resolved by the implementation of personalised medicine approaches. The aim of this project is to use patient-derived xenograft (PDX) models of neuroblastoma to assess whether therapies selected through molecular profiling of tumours and unbiased examination of ex vivo drug sensitivity can provide improved responses over that seen with conventional chemotherapy.

Methods: PDX models of neuroblastoma established from PB/BM at disease progression or post mortem and their matching cell lines were obtained from Children’s Oncology Group (USA). PDX models were expanded by subcutaneous implantation in NSG mice, and processed for serial engraftment, cryopreservation, and molecular profiling. Gene expression profiles were generated for the PDX tumours and matching cell lines, and used to select targeted therapy for subsequent testing in the PDX models. Optimal seeding densities were determined for cell lines prior to screening with the NCI-Approved Oncology Drugs set.

Results: The engraftment rates were high (90–100%) and tumours reached 1 cm3 within 6–8 weeks. Fluidigm analyses on 96 genes commonly up-regulated in cancer confirmed similar gene expression profiles between xenograft tissues and cell lines, and revealed potential therapeutic targets for each xenograft. Results from drug screening demonstrated varied anti-cancer activities of chemotherapeutic agents dependent on the cell line.

Future directions: We will use the PDX models to compare conventional therapy to agents chosen based on profiling and screening. Tumour samples are being collected from neuroblastoma patients treated within the Sydney Children’s Hospital Network to expand the number of PDX models available.