CSIRO fields cancer diagnosis tools

CSIRO Bioinformatics, a spin-off company from CSIRO Mathematical and Information Sciences, has developed new statistical methods it says could lead to improved diagnosis and treatment of cancers.

The new technique, currently being marketed to potential pharma and biotech partners, analyses the activity of genes in cancer cells and potentially allows the identification of the identification of the cancer type.

"We have done a couple of global trips already, and spoken to some 30 or 40 companies, out of which we are in serious dialogue with three or four," said Dr David Mitchell, Business Development Manager for CSIRO Bioinformatics.

"They include top 10 drug and diagnostics companies so we are pretty pleased with that. We are clearly in the top two or three in the world with our technology,"

Mitchell said the race was on to find cancer diagnostics, and the CSIRO Bioinformatics' statistical data mining tools helped identify the most relevant information from the large amount of information generated by new screening techniques.

The fact that many hundreds of genes may be involved, some of them working together, makes it difficult to develop a simple test to distinguish between different sorts of cancer, but the new CSIRO approach mines this data, enabling a simpler view to be gained of the cells. The technology is able to pinpoint a much smaller number of genes than the 50-100 or so normally identified by competing technologies.

"Everybody is in the race to develop accurate classifying tests in the next two to three years, but right now they are swamped with data. Our two-gene system is a major leap forward," claimed Mitchell.

The new system analyses the complex differences between cells, and identifies just one or two genes that are typical of different cancer strains. Mitchell described it as "literally like finding a needle in a haystack."

"Because only a few genes need to be looked at to tell what sort of cancer a patient has, pharmaceutical companies will be able to design simple, effective test kits to provide a detailed diagnosis," he explained.

The company is also currently investigating the potential of the technique for use in other areas such as diagnostic marker discovery, toxicogenomical marker identification, and pharmacogenomics.

"All three of these markets could be huge for us, and we know our technology could provide a real advantage in all three areas," said Mitchell.