by ecancer reporter Clare Sansom

The term “personalised medicine” is used to refer to the model of medicine in which treatment is tailored to an individual patient as well as to his or her disease: essentially, to get “the right treatment to the right patient at the right time”.

Although this model can be applied to any medical specialty, it is currently most highly developed and most often used in oncology.

Personalised cancer medicine is based on the assumption that, on the molecular level at least, no two tumours are the same.

Drugs can now be developed to target proteins that are expressed in a minority of cancers of certain types.

Oncologists are able to select certain drugs based on the genetic profile of a patient’s tumour, and this approach will be extended to cover many more cancer types, genetic markers and drugs in the near future.

It is therefore not surprising that the life science networking organisation One Nucleus recently chose personalised medicine as the topic for one of its Life Science Leadership Seminars, or that many of the delegates at this seminar were involved in basic or clinical cancer research.

One Nucleus was formed in 2010 by the merger of two smaller organisations based in London and Cambridge and now serves an area that includes about 60% of the UK’s life sciences industry.

It holds three or four leadership seminars a year; each covers either an aspect of technology or a therapeutic area and attracts a wide range of delegates from both academia and industry.

This meeting was held at Chesterford Research Park near Cambridge in mid June 2014.

The first speaker, Professor Dame Janet Thornton, director of the European Bioinformatics Institute (EBI) in neighbouring Hinxton, presented the scientific background to personalised medicine and therefore set the translational research discussed later in the day into context.

Thornton explained how the cost of sequencing a human or equivalent genome has dropped in just a decade from several billion dollars to about a thousand, perhaps equivalent to “the cost of an annual Arsenal season ticket”.
  
It is now possible to sequence the genomes of many individuals fairly quickly and easily, and a number of national and international projects have been set up to investigate human variation by sequencing multiple genomes.

In the UK, the Department of Health has set up the 1000,000 Genomes project to sequence that number of complete genomes from NHS patients before the end of 2017.

Thousands of these patients will have cancer, and each cancer patient will contribute two genomes to the project: one from the tumour and the other from normal tissue.

Cancer is one of three clinical foci for the project, along with rare diseases (mainly monogenic disorders) and infectious disease.

This project alone will generate more human gene sequence data than has yet been obtained, and there are many as yet unsolved challenges associated with storing, maintaining and analysing this volume of data.

It will be necessary to combine basic biological data with clinical data derived from individual patients.

This brings up issues that basic biomedical scientists have rarely had to deal with, ranging from ethical and legal safeguards to the fact that clinical data is held in non-standard formats that vary between countries and sometimes even between hospitals.

Thornton ended her talk by introducing the new Centre for Therapeutic Target Validation which was launched at Hinxton in late March 2014.

This is a public-private partnership involving the EBI, the Sanger Institute and pharmaceutical giant GSK, and aims to develop a bioinformatics based framework for selecting and validating protein targets for drugs against a range of diseases including cancer.  

All the data generated in the new centre will be made publicly available, and research will be published promptly in peer reviewed journals.

This tour de force of an introduction was followed by Alan Wookey of clinical trials company LabCorp, who discussed clinical trials for companion diagnostics.

A true personalised medicine is only of use if it is registered alongside a test – most often a genetic test – that can be used to distinguish between patients who are likely to benefit from the therapy concerned and those who are not.

One of the earliest examples of combining a drug with a diagnostic tool was AstraZeneca’s oncology drug gefitinib, which can only be prescribed for lung cancer if a test has indicated that the tumour carries mutations in the gene EGFR.

Another well-known example in oncology is the BRAF kinase inhibitor vemurafenib, which is only effective for malignant melanoma that carries the V600E mutation in that kinase, and which is now marketed with a laboratory test for the mutation.

It is now essential to develop a targeted therapy and its diagnostic test together and if at all possible to submit them for approval at the same time, remembering that a diagnostic is only of use if it has the potential to change clinical practice.

This is a complex area and many pitfalls remain for the validation of diagnostics and for the design of clinical trials to test combinations of diagnostic and therapy.

However, about 15 of these companion diagnostic tests are already listed on the FDA website, almost all for oncology indications, and many others – including some for diseases other than cancer – are in the pipeline.

The challenge of selecting the right partners in developing personalised medicines, and the right patients for clinical trials, was addressed by Dimitar Tonev from HCV Research UK.

HCV Research UK is a consortium funded by the Medical Research Council that brings together clinicians and scientists throughout the country working on the chronic liver disease hepatitis C, which is one of the most important liver cancer risk factors.

The consortium holds data on and samples from a large cohort of about 10,000 HCV-infected patients at all stages of the disease.

The database and biobank are maintained as resources for the research and clinical HCV communities in the UK.

Currently, 71.2% of the patients in the database are male, most are between 40 and 60 years of age and the most important route of infection is the injection of illegal drugs, often many years before the first symptoms occur.

Interestingly, white Caucasians are more likely to be infected with the type 1 genotype virus and Asians with the type 3 genotype virus.

This data is being used in clinical trials of new drugs for HCV and investigations into prognostic factors, including risk factors associated with hepatic cancer.

Grace Macaulay of Medimmune, the biologics arm of AstraZeneca, gave an engaging presentation about the risks and rewards of running clinical trials in emerging regions, with a particular focus on Africa.

This is important in personalised medicine because disease stratification means that clinical trials need to recruit large numbers of patients, sometimes with a variety of genotypes or other characteristics, in order to stand a chance of reaching a statistically significant result.

This often means that it is necessary to recruit patients internationally, and to set up local clinical trial centres in all the countries involved.

Furthermore, the burden of chronic disease including cancer is predicted to shift away from the West; about 70% of deaths from cancer already occur in low and middle income countries.

Macaulay was optimistic about the opportunities that African countries, and particularly her native Nigeria, can offer to companies involved in setting up clinical trials.

She presented a SWOT analysis of these, identifying a huge, growing and mainly young population and a growing, well-educated middle class among the strengths and opportunities.

Weaknesses and threats include a relatively low number of physicians for the population, limited resources for research and public health infrastructure, and a lack of experience and training.

The American Society of Clinical Oncology (ASCO) is helping to address the last of these by organising clinical trials workshops in Southern countries; Macaulay ran one of these courses in Abuja, Nigeria in 2012.

Personalised medicines and companion diagnostics will be of very little use until or unless their utility has been recognised by the prescribing physicians.

Mark Bartlett, managing director of a London-based startup genomics company, Geneix, presented a compelling case for improved communication between pharmacogenetic data professionals and clinicians.

Genomic data and the results of complex diagnostic tests need to be made available to clinically appropriate timescales and in a format that a busy doctor can immediately take on board.

He made two suggestions for improving the interpretation of the avalanche of available data and the advice given to clinicians: improved digital communications and, as previously proposed by Thornton, the introduction of a new profession of ‘genomicist’.

This is bound to become even more important as the number of clinically relevant, validated biomarkers and the volume of data increases,

In oncology, we already know of perhaps 20,000 different tumour markers, but only 20% of cancer patients can currently be treated with a targeted therapy that has been linked to one of these.

Researchers at Ariana Pharma, a spin-off from the Pasteur Institute based in Paris, are developing robust biomarker signatures that can select appropriate therapies for at least some of the other 80%.

Mohammad Afshar from Ariana described their trademark data mining tool OncoKEM® which can rank a set of possible treatments for an individual cancer patient based on the "genetic distance" between tumour and normal tissue, as measured by multiple technologies including NGS, gene expression and microRNA profiling.

This tool has been used in clinical trials and is gaining acceptance in the oncology community.

Ariana is now collaborating with scientific instrument company Bruker, combining biomarker identification with NMR spectroscopy to develop a tool for distinguishing brain tumours from the surrounding normal tissue in “real time” during surgery.

Other presentations at the meeting included one from Philip Oliver and Dona Finch at Medimmune on patient stratification for respiratory disease, and discussions of the regulatory procedure for personalised medicine and its relevance to rare diseases.