How AstraZeneca achieved a fivefold increase in its new medicines success rate
The launch of a Functional Genomics Centre in collaboration with Cancer Research UK represents the latest stage in a driving ambition from AstraZeneca to improve its hit rate.
Developing new medicines is a notoriously difficult and long-winded process, ending in failure far more often than success.
Briefing the Cambridge Independent, AstraZeneca’s Mene Pangalos recalled how a review of drug projects led to a cultural shift in the company’s approach that generated a major improvement in success rate.
“As an industry, we are very good at failing,” said Dr Pangalos, displaying a graph that showed that between 2005-10, the proportion of molecules entering the pre-clinical setting that went on to pass phase III clinical trials and be launched as new medicines was a paltry six per cent.
“At AstraZeneca we were even better at failing – we had a four per cent success rate,” he said, noting that the company was incentivised “on the quantity of what it was doing – the number of candidates, the number of R&Ds – not on the output, which is medicines”.
Indeed, AstraZeneca was the second most productive pharma company after Pfizer during the period in terms of the pipeline volume.
Dr Pangalos, who joined the company in 2010, added: “Our incentives really weren’t aligned with what our job is, which is to turn science into medicines and translate that to patients.
“We looked at what we’d been doing and other companies and came up with a framework that we thought would improve the probability of success for our programmes. We cut back the pipeline significantly and really focused on the quality of our research and development not on the quality or volume.”
Dr Pangalos and colleagues published a paper in 2014 in Nature Reviews, exploring the causes of drug project failure and identifying the five Rs for successful R&D programmes – the right target, the right patient, the right tissue, the right safety and the right commercial potential. Together, they implied a sixth R – the right culture, for the review changed how AstraZeneca operated by demanding its scientists ask key questions before progressing a drug candidate.
“How well do you understand the biology of the target you are working on and the disease you are trying to treat? How do you try prove or disprove your hypothesis? If you have a molecule, let’s say it inhibits a kinase in a tumour, do you understand the level of inhibition you need in that tumour and the chronicity of that inhibition you need? Do you need to inhibit the enzyme 100 per cent for 24 hours, or 50 per cent for 12 hours?
“Understanding that pharmacodynamic relationship is incredibly important,” said Dr Pangalos.
“As a result, none of our programmes goes into the clinic now without being able to demonstrate proof of mechanism, that you are engaging the target, the pathway, so you know that if it fails in the clinic, you’ve tested your scientific hypothesis and you have learnt something. We had way too many examples of us failing in the clinic and when we’d ask the question ‘Did we test the hypothesis?’, there were quizzical, blank stares. We don’t do that anymore.”
The approach to the ‘right safety’ was key because the company’s focus on volume from 2005-10 meant it “was very good at finding ways of getting around the safety signals, where you are not worrying about proof of mechanism”. By lowering doses, molecules were able to enter the clinic, but proved ineffective at engaging pathways.
Meanwhile, defining the right patient population “most likely to respond to that particular drug or therapy” is now considered essential.
“If it doesn’t work in that specific patient population, it’s not going to work in a broader population, particularly important in oncology but actually it’s important in all of our disease areas.
“Today, more than 90 per cent of our pipeline has got a precision medicine strategy – across diabetes, respiratory and cancer,” said Dr Pangalos.
The final R – right commercial opportunity – demands that there is a proper understanding of the value proposition, and why someone would reimburse and pay for the medicine.
“It really helped change and drive behaviour,” said Dr Pangalos.
A second paper, published earlier this year, showed that AstraZeneca’s success for 2012-16 had shot up to 19 per cent: a nearly fivefold increase.
And this figure does not include the pipeline for AstraZeneca’s global biologics arm, MedImmune, which can boast a slightly higher hit rate.
“It’s in the top three pharma performance,” said Dr Pangalos.
But he added: “We’re failing now 80 per cent of the time. We’re still pretty good at failing. How do we get this even better? This still isn’t good enough for us.
“We’re not complacent. We’ve had a good run and our numbers are staying a little bit above this. But we want to make sure that, unlike any other company, we keep this going not just for five years, but 10 years, 15 year, 20 years.”
Today, the main reason for failure in AstraZeneca’s drug candidates is efficacy.
“We’re now testing the scientific hypothesis, but our science is wrong,” said Dr Pangalos. “That’s really, really important.”
Indeed, in December AstraZeneca faced the news that its combination of Imfinzi and tremelimumab had not improved overall survival versus standard chemotherapy in metastatic head and neck cancer previously treated with platinum-based chemotherapy.
“What we’ve been trying to do over the past few years is broaden the therapeutic platforms that we can work in to enable our science to work on any drug target in the genome,” said Dr Pangalos. “I don’t want them to be limited by having to be a small molecule or an antibody.
Oligonucleotides, modified RNAs, CRISPR-based therapeutics, bicycles… we have a whole array of different platforms that expand the toolbox and enable the genome to be druggable.”
Dr Pangalos said the company had also responded by building “a much stronger genomics capability that enable us to identify rare variants, genetic variations in the population that enable us to stratify patient populations across all our disease areas”.
And the company is using artificial intelligence both in its biology and chemistry, which it will be expanding.
“Once you’ve identified variants in the genome, how do you work out the consequences in terms of their function on cells or disease?
“This collaboration enables us to do that with much higher throughput and much greater capability,” said Dr Pangalos, who is executive vice president of the company's new biopharmaceuticals unit, having previously headed up its Innovative Medicines and Early Development (IMED) Biotech Unit at the company.
In particular, the centre will enable them to develop the use of powerful CRISPR gene-editing techniques.
“CRISPR is a very powerful technology from a research perspective,” said Dr Pangalos. “It enables us to do things in terms of modulating genes in an exquisitely selective way.”
He explained this could involve knocking genes out – a process that used to take months but which can now be done in weeks – or repressing them, downregulating gene expressions.
Equally, CRISPR also enables the activation of genes.
“Sometimes we want to switch things on to stimulate the pathway in the right way. Having this capability of all three things is quite unique,” said Dr Pangalos.
And it hopes the Functional Genomics Centre will give it further industry-leading capabilities.
In five years, Dr Pangalos expects “more drug discovery programmes up and running and accelerating towards the clinic” as a result of the centre.
“What is also important is to embed the technology and the culture of functional genomics into everything we do beyond oncology. We want this to become routine,” he said.
Cancer Research UK Cambridge Institute director Greg Hannon says Functional Genomics Centre will help in drive to improve survival rates
Cancer Research UK believes the collaboration with AstraZeneca will be of great benefit to its researchers.
CRUK Cambridge Institute’s director Greg Hannon, who is originally from the US, told how a discussion with Milner Therapeutics Institute’s director Tony Kouzarides prompted the idea of a new genomics centre.
“Shortly after I arrived Tony showed up in my office and started to talk about his excitement around the Milner Institute. It probably wasn’t long after that that we started to discuss the possibility of creating what I termed a functional genomics centre of excellence for the UK, as a flagship programme in the Milner.
“I have to say it’s tremendously exciting four or so years after those initial conversations to see this realised as a partnership with AstraZeneca.
“The vision for this was really to try to bring together expertise from AZ and CRUK – both in functional genomics and broad biology that CRUK can bring to the table – and creating a world-leading centre of expertise. We refer to it as functional genomics but what it really is is genetics.”
Prof Hannon said the combined effort to drive forward technology would be crucial in helping the charity meet its goal of improving survival rates for cancer patients
“In the last 40-odd years we’ve seen about a doubling of survival rates,” he said. “The explicit CRUK goal is to move from two in four surviving to three in four in the next 20 years, but one of the difficulties is, as we move along this time continuum, we are dealing with more and more difficult cancers.
“To meet those goals, we need more and more sophisticated tools and a much deeper understanding of the biology and that’s really what this collaboration is meant to foster.”