ICR Chief Executive Professor Alan Ashworth hosted a series of presentations on personalised medicine. Personalised or stratified medicine matches the treatments patients receive to the genes driving their particular cancer type.
“Patient stratification using molecular genetic tests is increasingly used to define subgroups who might receive benefit from particular therapies. This session will highlight recent progress in this important area,” Professor Ashworth says.
The session also included a presentation from David Gonzalez de Castro from the ICR and The Royal Marsden Hospital
The challenges for the implementation of molecular diagnostics in the clinical setting
Reliable and cheap predictive tests must be developed and used in conjunction with new personalised medicines to ensure they are only given to patients who will benefit from them, Head of Molecular Diagnostics at the ICR and The Royal Marsden Dr David Gonzalez de Castro said.
However, developing these diagnostic tests alongside drugs had proven to be more complex than originally predicted and only a handful were already used routinely in clinical practice. Speaking at the NCRI cancer conference, Dr Gonzalez de Castro outlined a number of the challenges that scientists and clinicians face as patient care moves towards “stratified medicine”.
When designing these genetic medicines and tests, scientists had to “know your enemies as well as you friends” – they not only had to make sure they were targeting the correct gene that was causing the disease, but also had to understand other genes and biological processes involved in cancer development that may interact in unforeseen ways. They also needed to consider that tumours continue to “evolve” as they grow and spread, and may develop new mutations that render them resistant to treatments. Patients’ tumours would therefore need to be re-tested at stages beyond initial diagnosis.
For patient stratification to be routine and successful on a national level, tests would have to be “reliable” and “relatively cheap”, Dr Gonzalez de Castro said. In the US, the Food and Drug Administration was approving standardised diagnostic tests alongside new targeted therapies, however in Europe this was not the case. It was therefore important to make sure labs followed a standard process and continued to reassess, or “validate”, their own tests to ensure accuracy. Results would also need to be generated “very quickly”, so doctors could use them to make decisions on patient care, as any delays could hold up treatment. For this reason, patient samples should be “panel tested” with a variety of diagnostics. “You can’t do one, wait for the results and then do another, when patients are just getting worse,” he said.
Finally, although there was a general move towards stratified medicine, Dr Gonalez de Castro said it was important to “put it into context” of existing patient care. New targeted treatments may not always be the best option for patients, if old-style drugs are effective and cheap.“We need to work with clinical colleagues and make sure we have a sensible approach to treating patients,” he said.
Interview with Dr David Gonzalez De Castro for eCancer TV
Watch Dr David Gonzalez De Castro speaking with Professor Gordon McVie, about the key procedures of molecular diagnosis.
Listen to a podcast from the NCRI conference on the future of personalised medicine, with CRUK Director of Stratified Medicine James Peach and the ICR's David Gonzalez de Castro
Proffered Paper Session: Diagnosis and therapy
Professor Ros Eeles from the ICR and The Royal Marsden hosted a session that included a range of presentations about new research on cancer genes and new drugs.
The session included a talk by Professor Eeles on a discovery led by her team of seven new variants of the human genome that increase men’s risk of developing prostate cancer, while the ICR’s Farah Rehman discussed new treatment strategies for a type of drug called PARP inhibitors.
The world’s largest study comparing the genes of men diagnosed with prostate cancer with healthy controls has now found 40 regions of the genome linked to the disease, the NCRI cancer conference in Liverpool heard today.
Lead investigator Professor Rosalind Eeles gave an update of the PRACTICAL study, an international collaboration using a technique called a genome wide association study to look for genetic variants linked to prostate cancer.
The most recent stage of the project involved scanning the genome of 4,574 men diagnosed with clinically significant prostate cancer and 4,164 controls. The scientists found seven new genetic variants that were significantly more common in the prostate cancer cases then controls.
“We now have more than 40 variants identified that are associated with prostate cancer development… Importantly, they act multiplicatively,” Professor Eeles said.
Professor Eeles told the conference attendees that if you scanned the genome of all the men in the room and took the one per cent at highest risk, they would have four times more chance of developing prostate cancer than the average person. The top ten per cent would have a 2.4-fold risk compared to the population average. Professor Eeles said the number of genetic variants found so far had approached a point in which the information maybe useful for targeted screening and prevention programmes.
“For each genetic variant the increased risk is very small, but when you combine them together it’s very large. So we are now starting to find the kind of risks that could be clinically significant,” she said.
Developing rational drug combination strategies for PARP inhibitors
Dr Farah Rehman from the Breakthrough Breast Cancer Research Centre at the ICR discussed investigations into combining a new class of drugs called PARP inhibitors with other types of drugs.
ICR scientists have found drugs that may boost the effectiveness of a class of cancer drugs called PARP inhibitors, by screening a library of thousands of possible compounds. PARP inhibitors, including a drug called olaparib that the ICR helped to develop, are a new type of experimental drug that works by targeting weaknesses in some cancer cells. Cancers that have faults in their BRCA genes are known to be particularly sensitive to PARP inhibitors, and these are already showing success in trials of patients who have breast and ovarian cancer linked to BRCA mutations. The drugs selectively target the mutated cells, leaving other – normal – cells relatively unscathed. But when attempts have been made to treat patients with cancer not linked to BRCA mutations, the results have been mixed, ICR scientist Dr Farah Rehman told the NCRI cancer conference.
Similarly, efforts to combine PARP inhibitors with other drugs have been “a bit of a mixed bag”, including trials that had to be cut short because patient side-effects were considered too high. Dr Rehman said that the ICR team was taking a systematic approach to looking for cancer types which may be sensitive to PARP inhibitors, and drugs that may work in combination. “We wanted to look at drugs that haven’t been looked at before, and see if we can get around some of the toxicity we’ve seen,” she said.
The team screened 3,000 compounds and found several that were more effective in combination with PARP inhibitors. They then decided to try them in cancers linked to PTEN mutations. “We thought this was a good place to start, with something that’s known to be partially sensitive to PARP inhibitors, trying to make it more sensitive,” Dr Rehman said. Some drug combinations did seem to be selectively targeting the PTEN mutations, which made it less likely toxicity would be an issue. Dr Rehman said further work would focus on identifying the mechanisms underlying the interactions between the compounds and the mutation, “with a view to rapidly moving the combinations into clinical trial”.
Parallel Session: Integrating knowledge for the new age: cancer informatics
The science of information processing, called informatics, was becoming increasingly important in the field of cancer research, the ICR’s Dr Bissan Al-Lazikani told the conference.
Dr Al-Lazikani said new technologies such as next generation sequencing combined with the rise in large-scale international collaborations and genetic, molecular and chemical screening programmes was leading to a massive increase in data generated.
The field had become a “complicated beast” that included a breadth of techniques such as “data modeling, data integration, computational chemistry, network informatics and more”.
“This is all trying to really address this embarrassment of riches that we are all now facing; this vast amount of data,” she said.
“This growth in data is not going to help us become more productive without us becoming really smart about how to process and use it – rather than just put it in banks.” Dr Al-Lazikani said the ICR was addressing some of its data integration issues by launching a database called canSAR, which “tries to integrate data in a way that is useful to researchers”.
Parallel Session: Reducing the burden of breast radiotherapy
Tests are being developed to identify women who will not benefit from radiotherapy, so they can be spared the side-effects of treatment, Professor John Yarnold told the conference.
Many women with breast cancer are treated with radiotherapy to try to prevent the cancer coming back. Professor Yarnold said studies had shown that radiotherapy prevents 20 relapses for every 100 women treated, five of whom would have died from their cancer spreading. However, the other 80 women did not benefit from radiotherapy and yet still were at risk of side-effects from treatment, which can include breast shrinkage and hardness, stroke and cardiac injury. Prognostic tests were therefore needed to identify women who have a lower chance of cancer recurrence, he said.
Some studies had indicated that particular sub-types of breast cancer have differing risk of relapse following radiotherapy, however these were not yet conclusive.
These tests were “emerging but not established”, he said. “It’s interesting and promising but it’s still not something you would put into everyday practice. The traditional factors we use to classify and treat our patients are still at least as good as the molecular assays – at the present time.”
Other tests could be used to help determine the dose of radiotherapy given. For example, predictive tests were being developed that identify women at higher risk of recurrence, so they can be treated with a higher dose of radiation. There was also evidence that some breast cancer sub-types were more sensitive to radiotherapy than others, so in future the dose of radiotherapy could be matched to the tumour type.
Finally, some women experienced worse side-effects from treatment than others, and it had been suggested that their genes were playing a role. International studies were underway that were scanning women’s genomes to look for genetic variants that may predispose them to an adverse reaction to radiotherapy. A test incorporating these genes could therefore be used to adjust the dose of radiotherapy more closely to the patient, Professor Yarnold said.
The cost of caring for breast cancer patients in the UK is rising ten per cent each year, the conference heard, so there was an economic as well as a patient benefit to giving treatments to only those who would benefit. The current levels of £6 billion a year - £21,000 per new case diagnosed - represented six per cent of all healthcare costs.