Test predicting effective treatments from routine cancer samples approved

A new AI-based test that can predict the most effective treatment from images of routine cancer samples has been approved for use in the UK and EU, speeding up diagnosis and reducing the need for lab testing.  

Developed by Cambridge-based company Panakeia, the PANProfiler test analyses digital images of routinely collected breast tumour samples that are normally checked down a microscope by a trained pathologist to determine the presence of cancer.   

The usual next step would be to send a further sample for lab testing to identify the best treatment approach, with the results taking days or weeks and costing hundreds or even thousands of pounds depending on the test.

However, the PANProfiler Breast test skips the need for testing by directly predicting whether the cancer contains ER or PR receptors, marking the patient out as a candidate for hormone therapy, or HER2, targeted by the drug Herceptin.

All this happens from the original digital image in a matter of minutes, with accuracy comparable to lab testing, making the PANProfiler test far faster and significantly cheaper than existing tests. Not only does this save precious time in the patient journey but also significantly reduces the burden on busy and costly laboratory services, which are facing a backlog due to the impact of COVID-19 on cancer diagnosis.  

PANProfiler integrates into digital cancer pathology procedures and is currently being trialled in hospitals in the UK, with plans in place to expand into Europe, North America and Asia. As of 13th October, the test now has UKCA and CE approval for clinical use by health services in the UK and EU.   

Panakeia’s technology grew out of research by co-founders Pahini Pandya, a former cancer scientist at the University of Cambridge, and AI researcher Pandu Raharja-Liu. They realised that tiny differences in the appearance of cancer cells, which can only be detected by a computer, can reveal important information about their molecular state and the likely best treatment options.

Following the launch of PANProfiler Breast, the Panakeia team is developing similar tests for other tumour types.

Panakeia’s mission to speed up decision-making in cancer diagnosis and treatment is driven by Pandya’s own experience of waiting for the results of tests for blood cancer - the same disease that she lost her childhood best friend to - which fortunately turned out to be negative.

She said: “I know first-hand the anxiety of waiting for your test results. Due to the pressure on labs, even in the best healthcare systems, diagnosis and treatment decisions can take weeks - an unacceptable and stressful delay when dealing with a fast-growing cancer. We’re excited to be rolling out PANProfiler to hospitals here in the UK and around the world to speed up access to treatment and help save lives.”

Raharja-Liu, who has lost family members to the disease, added: “This is a golden opportunity to transform cancer diagnosis. We can now do something that nobody has achieved before - to see more from every tumour sample, gathering rich information about what these cells are like and how best to treat them.”

Professor David Harrison, director of iCAIRD, one of five centres of excellence in the UK focussed on AI applications in pathology and radiology funded by Innovate UK as part of the government's Industrial Strategy Challenge Fund, said: “This profiler has the potential to help the NHS improve efficiency in pathology laboratories, support efforts to ease pathology workloads, and facilitate COVID-19 recovery. It enables rapid cancer diagnosis, improving both patient experience and outcomes.”