The age of personalised medicines might not have quite arrived for patients but for companies in the business of developing and selling drugs, dawn has certainly broken. The huge potential seen in the ability to match the right drugs with the right patients – to improve R&D productivity and help gain regulatory approvals and reimbursement – has prompted huge efforts to be deployed across the industry to achieve these ends.
In the case of AstraZeneca, a personalised healthcare and biomarkers division was created last year in an attempt to infuse the tenets of targeted drug discovery throughout its research and development activities. Having already beaten a goal of having a diagnostic or patient stratification tool in place for half of its pipeline by 2014, Bob Holland, head of Astra's new department, tells EP Vantagethat he expects the focus on personalised medicine to transform drug development in the long term. With targeted drugs addressing conditions as diverse as psychiatry, inflammation and asthma already in phase II, the fruits of Astra’s personalised healthcare push should start to emerge soon.
With 100 full time employees – from molecular biologists, diagnostic and imaging specialists to mathematicians – a member of Mr Holland’s team is now embedded in every research project. Even before a candidate has been selected, when a particular target or mechanism is being discussed, how particular patients will be detected is a crucial consideration, he says.
“Right from the very inception of a discovery project we expect the scientists not just to create a new molecule, but also to be investigating how they will identify patients that will respond. It infuses the governance processes we have, and led to the creation of my group, which tries to take the inventions a discovery scientist makes and mount them in clinically useful platforms that can be deployed in large scale clinical trials, and become part of normal clinical practice,” he says.
As part of the company’s R&D shake up last year, AstraZeneca committed $200m to developing four main areas: predictive science, payer evidence, personalised healthcare and clinical trial design and interpretation. Mr Holland says a big proportion of these funds will be flowing into personalised medicine.
Unlike some of its peers, AstraZeneca has not acquired diagnostics capabilities, preferring instead to partner with independent specialists.
Very early stage work is conducted in collaboration with academic institutions or smaller companies with specialised know-how, to find new ways to identify patients who will respond to a new therapy. Once a biomarker or stratification technique has been identified, the new assay is progressed internally and when validated a larger more commercially focused partner is brought in, to develop and test the prototype in larger phase IIb studies.
“The one thing I’ve tried to hold fast with is that we start phase III with a prototype that can be commercialised by the diagnostic partner. To do that we have to work with them during phase IIb,” Mr Holland says.
And while personalised medicine frequently refers to complex diagnostic tools, such as identifying molecular markers of particular tumours, Astra is “agnostic” about which techniques it will employ to stratify medicines and patients. In some disease areas mathematical techniques will be developed to refine simple patient questionnaires, to improve diagnosis of a particular disease.
Astra has broad partnerships in place with the acquisitive Dutch diagnostics group Qiagen – which helped develop the diagnostic tool currently helping to revive once-flunked lung cancer drug Iressa – and Dako Denmark. In January last year a collaboration with the Danish firm was forged to develop companion diagnostic tests for multiple oncology projects, including biologics and small molecules.
The recent scramble for diagnostic assets means few large independent players remain and much expertise has gone in-house at big pharma, leaving the prospect of Astra one day having to collaborate with a rival.
“We’re not afraid of that scenario,” Mr Holland says, adding that Astra has no immediate ambition to buy in diagnostic expertise.
“We have a unique strategy of pursuing a small number of strategic partnerships with global players - and those partners are genuine strategic partners. Dako scientists are working alongside our scientists, in a way which is completely the opposite to a fee-for-service arrangement,” he says.
“That is our deliberate strategy which at the moment seems to be working.”
Of course real proof that the strategy is working will only come once robust phase III data emerges on successful “stratified medicines”. While Astra has a number of potential projects in mid and late-stage development, Mr Holland says these are still largely under wraps, until intellectual property can be put in place.
He stresses though that developments are much more than a pipe dream still lingering in the research labs. And that projects already exist in therapy areas beyond the most successful realm of personalised medicines to date – oncology.
“We have a psychiatry drug in phase IIb now which is following a patient stratification approach with a diagnostic that is truly a new invention,” he says, on which news could emerge in around 18 months.
“We have phase II drugs in inflammatory diseases, autoimmune diseases, asthma and neurology which are following this approach. We have another drug that is already on the market that we intend to extend into other indications using stratification markers that we are developing IP around,” he says.
Talks with payers have convinced Astra that this personalised push will help with reimbursement - a conclusion that most industry players already concur with. Mr Holland, like many others, is also convinced that these techniques will help address the R&D productivity failures that have plagued big pharma for the last few years, by improving the chance of success in late stage studies.
Regulators meanwhile are also trying to keep up with developments. Proposals released by the FDA in July in an attempt to define guidelines for the approval of companion diagnostics were broadly well received (FDA guidance reinforces importance of companion diagnostics, July 14, 2011). Mr Holland says AstraZeneca welcomed the proposals although the company will seek some refinement and clarifications – in particular on the approval pathways these tools must take.
Global harmonisation of approval processes is another area many are calling for, given the hugely different guidelines and rules in place in Europe, the US and Asia.
“There is that complexity that existed in the drug world 20 years ago and I think we’re looking to see how this evolves with time,” Mr Holland says.
The drive towards personalised medicine – to improve treatments, return on R&D and help justify the high cost of truly innovative drugs – is gathering speed. And while many companies are explicitly talking about their own efforts now, it seems likely new strategies being employed will gradually become implicit facts of drug development – the norm rather than novelty.
Research conducted by AstraZeneca concluded that while 10 years ago 3-4% of drugs could be described as stratified medicines, or medicines associated with a biomarker in their label, this now stands at 8%.
That figure is only going in one direction.
“This really is a radical transformation in how we are approaching R&D,” Mr Holland says. “Any transformational change normally takes a bit longer than anticipated, but in the end it often turns out to be more radical than people expect.”