Childhood cancer detectives embrace open-source AI solution

Hudson Institute

The old saying goes that a problem shared is a problem halved. In the world of cancer research, there are few problems bigger than paediatric solid and central nervous system (CNS) tumours. These tumours are the number-one cause of childhood cancer-related deaths.

Limited drug development for childhood cancers

Exacerbating this problem, the pharmaceutical industry puts little emphasis on developing new drug treatments for kids’ cancers, due to their relatively small numbers compared to adult varieties.

As a comparison, the US Food and Drug Administration (FDA) approved more than 500 adult cancer drugs in the last four decades, but only 12 for childhood cancers, and none for brain or CNS cancers in kids.

“Paediatric cancers are unique from adult cancers, so we don’t necessarily expect the same drug that works in adult cancers to work in children,” said Dr Paul Daniel.

A global solution for children with cancer

Hudson Institute cancer specialists, leaders of the Victorian Paediatric Cancer Consortium (VPCC) Precision Medicine Program, realised that the more this problem is shared, the greater the chances of success.

Faced with such a monumental challenge, Hudson Institute’s Centre for Cancer Research, headed by Professor Ron Firestein, decided to share the burden, bringing highly competitive institutions together in pursuit of a common goal: to accelerate the search for new treatments and cures.

Creating the Childhood Cancer Model Atlas (CCMA)

Together, they created the Childhood Cancer Model Atlas (CCMA), a highly sophisticated paediatric cancer hub offering an open-source bank of childhood cancer tumour tissue samples complemented by advanced artificial intelligence (AI) data-mining tools.

With the world’s largest collection of high-risk paediatric cancer cell lines consolidated in one location, hospitals worldwide now collaborate by sharing tumour tissue samples with Hudson Institute researchers. These researchers utilise the samples for comprehensive testing against a wide array of cancer drugs.

Milestone in paediatric cancer research

That means researchers working anywhere in the world on childhood cancers can have their potential treatments tested then analysed using the latest AI techniques, with the results available for every paediatric oncologist and childhood cancer researcher worldwide.

The story of the CCMA featured in the esteemed journal, Cancer Cell, marking a milestone in paediatric cancer research and embedding Hudson Institute as a global leader in the field.

The role of bioinformatics in cancer research

This work goes way beyond the testing of tissue samples. Dr Claire Sun is a bioinformatician, whose job involves compiling, collating and condensing huge amounts of data, as the team performs hundreds of tests on each tumour type.

“Having models [samples] of various childhood cancers is the only way that we can study the disease and understand how it develops, what makes it grow and ultimately what treatments can be used to fight it off and hopefully cure it,” Dr Sun said.

/Public Release. View in full here.