The Pharmafocus Interview: John Patterson

pharmafile | December 19, 2007 | Feature | Research and Development, Sales and Marketing |  AstraZeneca, drug development, industry relations 

John Patterson is an AstraZeneca man through-and-through, having spent nearly all his working life at the company.

But over the course of his remarkable career in AstraZeneca, no role has held more importance than his current one – that of overseeing the regeneration of the company's drug pipeline and rehabilitating a reputation badly hit by a series of late-stage product failures.

Patterson has made it clear how he sees this will be achieved – by simultaneously improving the quality, cost and speed of AstraZeneca's drug development process.

He has also publicly committed himself and the company to achieving a highly ambitious target – to cut AstraZeneca's total development time for a drug to just eight years, down from the current industry average of around 10 years.

He told a recent analysts' meeting: "This is a management team that is tackling the industry in a way that I don't believe anyone else is. Just add up the number of major initiatives that are going on here, and I think that this is a company that is determined to succeed."

The regeneration job is clearly not yet finished – the company has made it clear that it is still looking to acquire extra phase III drugs to bolster its late-stage pipeline. But after nearly three years on the project, Patterson is keen to show the outside world how much progress AstraZeneca has made in re-balancing the riskiness of its portfolio, cutting development times and producing first-in-class and best-in-class compounds.

A lifetime's work

Joining in 1975 what was then the pharmaceutical division of ICI, Patterson was a newly trained medic and was given a leading role in the development of the breakthrough breast cancer medicine, tamoxifen.

Following the merger of Astra and Zeneca in 1999, Patterson took on a senior commercial role, and helped the company compete with the industry's biggest players in an era when sales and marketing were the key battlegrounds.

But in December 2004, John Patterson was called back into R&D when it became apparent to AstraZeneca's top executives that the company's late-stage pipeline was in deep trouble.

Over the course of a few years, a series of major drug failures decimated AstraZeneca's late-stage pipeline and knocked investor confidence in the company. The first casualty was lung cancer drug Iressa, which was eventually abandoned in December 2004 after a crucial study failed to show its benefits over placebo.

That same month, Patterson was drafted in to begin his review of AstraZeneca's processes, and the company also signed a deal with UK biotech company CAT to develop products. But these long-term measures couldn't help the company prevent further losses in its phase III pipeline.

After several years of doubt, anti-clotting treatment Exanta was finally withdrawn from all markets in February 2006, with diabetes drug Galida scrapped three months later. Cerovive, a first-in-class stroke drug followed in October. A further drop-out occurred in early 2007, when AstraZeneca abandoned AG-1067, its phase III atherosclerosis collaboration with AtheroGenics, in March.

With these failures still fresh in the memory, and no major launch since Crestor in 2003, some investors are yet to be convinced that AstraZeneca has turned the corner.

But Patterson says AstraZeneca is now able to show the world how far things have come in the last few years. Moreover, he says he, and everybody involved in R&D, is determined to prove wrong the critics who had written off the company's pipeline.

Why is now a good time to talk about research at AstraZeneca?

"As with any big organisation, change takes time and we are now beginning to see some of the changes we have made take effect," says Patterson. "Ultimately, what matters is the medicines we deliver, and we are really very proud of the way we are starting to re-create the pipeline, move it faster and deliver late-stage molecules and change the outlook for the business.

"We are now creating a 'bow wave' of products coming through the pipeline, but we are just a few years into a 10-year period in which I think all the benefits from the changes will emerge."

Speeding up the R&D process

AstraZeneca has gone back to basics, challenging some of the long-established practices of drug discovery and development.

Part of this process has been undoubtedly painful for some of the company's R&D staff, as around 700 positions will be cut worldwide, with reductions in data management, regulatory and other departments.

But beyond headcount reductions, Patterson's team has scrutinised all elements of R&D work, breaking down each stage into individual elements. Naturally enough, the team began with pre-clinical work.

"We looked at how long it took us to go from lead identification to lead optimisation to nominating a compound, and our discovery colleagues have done a huge amount to reduce that time."

In cardiovascular and gastrointestinal pre-clinical work, the company has cut down the lead optimisation process by a year, down from 30 to 18 months.

The early signs are encouraging. Just six months into 2007, AstraZeneca had put 14 new molecules into phase I studies, two more than for the whole of 2006, which was itself a record year.

Turning its attention to phase I, the company has also rapidly accelerated one-month toxicology studies, with the average time taken down from 9.3 months to 6.2 months.

Meanwhile, the time taken between the protocol being written to the first subjects being tested in phase I has been cut by three months in one key oncology project – a massive saving in time and money, which AZ hopes to apply across its portfolio.

Finally, phase II development time is now also being cut, with new technology allowing the company to conduct smaller studies and reduce patient recruitment time and cost.

From series to parallel

So, how has the company achieved these feats? In the pre-clinical stage, much of the time saved in the new system has been achieved by running tests in parallel where once they were conducted consecutively.

"You might have a molecule which you think is your lead molecule, and you want to test whether it has any potential for [the heart side-effect] QT prolongation, and it looks all right. Then you think 'we'd better do a bit of non-GLP toxicity work', then you do a bit of something else, and then you do your non-target clinical pharmacology testing. That all adds ups, and you say 'yes, it's clean, it looks OK'.

"Whereas now, we are saying – this looks like a really good molecule; let's do all of those things in parallel and get all the results back at the same time. That way, we can make a single decision about whether to continue with it or not, and that it is much faster."

So why hasn't this been done before?

"The answer is that you have to spend more money up front, you have to organise yourself to do it, and you have to synthesise enough of the molecule. There are many different reasons; sometimes, there are doubts about a molecule, so you think 'better not do that just now'.

"All of those technical reasons, together with human nature have created that sort of situation. The biggest example of that is between phase II and phase III. Classically, you do your phase II programme, the last patient is treated and you have a period of time where you get your data in and you have another period where you analyse it, then you say 'oh, that looks good', after which you go and talk to your management committee and they say 'yes, that's OK, carry on'.

"So, you then go and write your protocol and see the FDA and so on. You actually lose 11-18 months when not a single patient gets dosed," says Patterson.

"But if the phase IIa data looks really good, you can say, we're going to take a risk. That means having everything in place to start phase III very soon after phase II ends. If you do a campaign to synthesise more material, formulate it properly, write the protocols before you have the data from the first studies, you can do it in six weeks."

Mission Impossible

For anyone not involved in drug development, it is difficult to grasp just how difficult a challenge these delays and time-lags of conventional practice have been. One example is the time-lag of weeks or months between the regulator granting approval for testing in humans to begin (called the Investigation New Drug application or IND) and the trial actually commencing, which could be weeks or months.

"We set ourselves a target in oncology to be dosing our first studies within 48 hours of the IND being opened. Our team actually called it Mission Impossible, but they achieved that, which is a tremendous success."

Patterson says: "So, the question is why weren't we doing that before? Well, before when you had the IND, you wrote your protocol and that had to go through the ethics committee etc. This way, you write your protocol and you go to the ethics committee subject to the IND being issued. We get the legal agreement, which was taking nine months, written in advance.

"It is just about applying what most people would say is common sense, others would say it was a production line or industrialisation to what is a technical process. To me, what is exciting are the stages of development in designing the experiment and then the analysis of the data – the bit in the middle should be run as quickly and efficiently as possible.

Translational medicine and the Critical Path

'Translational medicine' (an umbrella term for a raft of emerging technologies) has become the buzzword in the industry in recent years because of its ability to give signs of success or failure for a drug before it reaches phase II or beyond.

Patterson says use of translational medicine and biomarkers is now embedded in AZ's processes.

"We now have a golden rule for every molecule going into clinical development. The team has to have considered what it can do in terms of biomarkers. It's particularly helpful in oncology, but not just confined to there.

"It is not just about having a biomarker; it's things like imaging studies. So, for instance, the key questions in neuroscience are does it get into the brain, and is it binding to the receptor?

"Imaging studies – for example, having the right kind of PET ligand – is an incredibly helpful tool. Of our last seven oncology products that have gone into man, six of them went in with a marker.

Classically, in cancer drug research, you would keep raising the dose until it produced unacceptable toxicity, which is why you do the phase I studies in late-stage cancer patients rather than healthy volunteers.

But with biomarkers, you can screen out drugs which are toxic before they reach a therapeutic dose, and which are therefore not worth developing. That is of benefit to the patients, obviously, and it is also of benefit to us because it allows us to go back to the drawing board and do something different.

The FDA launched its Critical Path Initiative in 2006 to help speed up drug development. Has that had an impact yet?

"I would say it hasn't had any impact so far, but it's an important initiative, and the people leading it at the FDA really do understand how regulatory medicine has to change.

"But I think translating [that approach] to the individual medicine as it is coming through is a much harder process, and I would say many reviewers in the depths of the FDA and the EMEA are still working on much more traditional lines."

The safety of drugs which are already on the market is of course a big issue in these post-Vioxx days. Would it be fair to say that pharmacovigilance systems need to move onto the next level of sophistication?

"I think post-marketing surveillance systems are pretty well developed, and they have served us well over the years, but there is a perception out there that they are not good enough, and it does come back to Vioxx.

"It will be interesting looking back at Vioxx in 10 years' time to see what the final verdict is, how the risk:benefit ratio really plays out. But at the moment, it's a cause celebre and there is a belief that the regulators are too close to the industry and have not being doing their job properly. They've been getting a lot of criticism for that, but in my view, pretty unreasonably.

"Don't forget that companies like us try to collect adverse events on a global basis, which is something that national authorities can't do.

"And though some people might not believe it, it is absolutely not in our interests to keep a medicine on the market that has a problem, or not to warn them about something that might be there.

"I give the challenge to my organisation that we should be the ones that pick up all the issues on our agent and we shouldn't be relying in any way on third parties."

Patterson says the development of electronic patient records will open up new possibilities in studying data on a large number of patients that hasn't existed previously.

"There's an opportunity to change the game and pool data from real life such as hospitalisation and major events that are occurring. That isn't currently readily available to us in the pharma industry, regulators or the health systems, but could be accessed if we all joined together."

Turning to perhaps best known of the AstraZeneca drugs which fell at a late stage, Exanta. Looking back, what went wrong?

"I guess journalists always want to know what went wrong or who was the guilty party – I would turn it round and answer you indirectly.

"Exanta was a first-in-class that could have revolutionised anti-coagulation – it had been 50 years since warfarin was discovered and nothing better had been developed.

"In short-term usage as a post-surgical DVT preventative agent, it worked beautifully and we had no problems at all at 10-11 days' dosing. In the phase II and phase III studies, we had a small number of patients who experienced changes in liver enzymes, but these appeared to be transient, even in patients continuing on the drug."

"However, a handful of patients (less than 10) went on to have a severe liver problem, and it became uncertain whether this side-effect could be detected or predicted, and whether it was an acceptable risk.

"When you consider that approximately half of patients who get warfarin don't tolerate it and consequently end up taking nothing, that means they are far more likely to die from a stroke than from any medicines you give them. So, it was about a risk:benefit ratio, and the belief in the company at the time was that that ratio was OK.

"We shared all our data with the regulator throughout the programme, and it was really only right at the very end that data emerged which did not show the benefit over warfarin that you would have hoped," says Patterson.

One shortcoming of the study design was that Exanta was compared with warfarin in a carefully controlled setting in which the INR test for blood clotting was carried out regularly.

"If you exclude all those people who don't tolerate warfarin and you carefully titrate their INR every week, then warfarin is a very good anti-coagulant and, so, beating it in those circumstances is almost impossible. But actually out in the wild, that doesn't happen – people have extra orange juice for breakfast, or don't go to the clinic for two or three weeks, and their INR goes all over the place, they get bleeds or they don't tolerate it.

"In the artificial circumstance in which we put it, the risk:benefit ratio appeared to be against the drug. That meant that the FDA, who did not appear concerned at all about these issues during our programme, when they did their final review, switched their view [and decided to reject Exanta.]"

He adds: "We would have carried on with the drug, I think, for the rest of the world, but something else then emerged. We extended the use of the drug in post-surgery in prophylaxis of DVT and pulmonary embolism to 30 days in a new trial that went on at a later stage because it was becoming commoner to do that in clinical practice, and we discovered a cohort of patients that got liver changes after the drug had stopped.

"Some weeks later, after stopping the drug, a small number of patients had a significant change in the liver enzymes, and that was the point where we couldn't continue. We could no longer predict who might get the problem and stop quickly if they got the problem. So that's why Exanta got stopped. "The issue was the interpretation of the risk:benefit ratio, the identification of the area of concern to the regulators, and what we might have done during the cause of that to mitigate it, but you know, hindsight is a wonderful thing.

"Everybody can look back and say knowing what I know now, I would have done something different. But what it did do was take us back into our whole portfolio and make very sure we identify the key issues and risks, and then decide if we can mitigate them or not.

Patterson said the first thing he did when he began his new role was to get drug development teams to draw up a 'risk list' for lead molecules.

"That risk list might include, say, liver enzyme changes, or maybe the drug is poorly absorbed or it has a low therapeutic ratio and the team says here is what we are doing about it," he explains.

AstraZeneca is now developing another anti-coagulant in the same class, and has learned the lessons from Exanta.

One problem the developers faced was that the new drug had a relatively short half-life, meaning its effects would peak then drop away quickly. To get round this problem the company has now developed a sustained-release formulation which gives a smooth, prolonged in-range plasma level that should avoid bleeds in the phase III programmes.

"This is the kind of development activity we hope will reduce the likelihood of having [late-stage] problems, so molecules can get through to the market-place as meaningful medicines.

So that drug is in the pipeline now?

"Absolutely, it is one of the projects in phase II, and hopefully we will see it move into phase III next year."

So, that could be a triumph of adversity in the end?

Yes, that would be nice. But of course, in a competitive field, there are a number of products. So, instead of being in the lead as we were with Exanta, there is a number of other drugs potentially ahead of us."

Over-reliant on innovative drugs

AstraZeneca's executives have already publicly acknowledged that its pipeline became too dependent on risky new first-in-class medicines, and re-balancing the risk has been a key priority for Patterson.

"We have a really proud history as a company of producing novel medicines, and we are not going to stop. But it is quite clear that you can't have 80-100% of your pipeline as first-in-class agents and expect to have a big enough hit-rate to sustain the business.

"So, we have been trying to get a better balance into the portfolio, to make sure we have a mixture of best- in-class and first-in-class. After all, we believe that Seroquel and Crestor are best-in-class, and they're are not bad drugs!

"It is clear that not aiming to be first to market can allow you to design a drug with improvements. For example, AZD 6140, our anti-platelet drug, we believe will be a best-in-class – better than clopidogrel, [Sanofi-Aventis' Plavix] and with a reversibility that allows huge advantages in the clinic.

"In future, you will see us with a mixture of both. If you look at our phase III pipeline today, we have some first-in-class cancer agents, and some that are best-in-class, so you achieve a reasonable balance.

Has this process been an evolution or a revolution?

"It has been a fundamental change to the way we do business, but it hasn't been based on turning the whole organisation upside down.

"I really don't believe you achieve change just by changing organisational structures. So, we didn't form fancy-named groups or anything like that, but we did put into place some absolutely key principles.

"We built a professional project management organisation, and we made leading projects an absolutely key job in the organisation. We have completely re-graded it, and put some of our best people into project management, which is revolutionary in many ways.

"We have changed our approach to doing things in many different ways. As I say, some people would advocate a revolution where everyone gets a new title and re-applies for their jobs; we did some of that, but not as a rule. It is about process and focus on delivery. It is about changing people's mindset.

"We have some very good people in this organisation, and we have delivered some very good drugs – they have been part of that in the past, but we had lost our way, lost our direction.

"My job was to get a team of people together who reinforced the direction we needed to be going in, communicated that well through the organisation, and then allowed all of those good people to start running with it.

"As I said, the most gratifying thing is how well they do that. And when they really understand what the leadership is asking them to deliver, they over-deliver, saying we can do this as well, and this, too."

What marks AstraZeneca out from other companies today?

"It is hard to say because I have never worked in other companies. It is an absolute determination to show the world that we are much better than they have been saying we are. Throughout the organisation, there's a belief that we can generate a pipeline of medicines that will make a difference to patients. So, there is a focus to improve our reputation, and restore it. That is about delivering medicines that make a difference.

"And the mantra throughout the organisation is about delivering 'Quality On Time' – quality projects, quality products, quality people. Delivering in a much faster manner.

"I have been saying that for about two years now, and talking about this triangle of values – quality, cost and speed, which has to be in everything we do.

Have people in AstraZeneca been stung by the criticism?

"Yes – our people read the papers and their families read the papers, and when it says AstraZeneca, the drug giant with the problem pipeline or with no pipeline, they don't like that.

"So, yes, they are stung by that and they know that AstraZeneca has won many awards over the years for being the company with the most medicines from its own research labs. But it has never been a smooth process – nobody delivers two drugs a year every year. You may get four drugs one year and none the next. That's because it is a series of experiments, it's not a simple timeline.

"So, they have been really stung and they are determined to show that with the right organisation, and the right leadership and the right products we can change the world."

Integrating MedImmune

In April this year, AstraZeneca announced that it was buying the US biotech company MedImmune for a record-breaking $15 billion.

Some analysts felt AstraZeneca had paid over the odds for the company, but AZ insists the deal will prove to be good value for money in the long term.

Patterson was involved throughout the MedImmune acquisition, from evaluating to pricing, and then integrating it into AstraZeneca and its existing biotech arm, Cambridge Antibody Technology.

Insisting that the deal is an "outstanding deal", Patterson says its gives AstraZeneca, capabilities that are "not available anywhere in any form in the market today".

The acquisition will allow the company to meet its target of biological candidates, making up one in four of the new molecules offered for phase III.

MedImmune also has the crucial manufacturing capacity, which Patterson says now makes AstraZeneca a "fully fledged bio-pharmaceutical company".

The MedImmune acquisition and the clutch of in-licensing deals the company has sealed in the last two years clearly represent a major shift in strategy for AstraZeneca, which had had relatively few external alliances before.

"When you have got a gap in your pipeline, you can't invent your way out of it. While we are doing a huge amount to pull through the early products and be more productive in discovery, that takes time.

"So one change we had to make to our mindset was, we are going to externalise and bring in projects to fill in the gap. But we are also doing that because there is a huge world of science out there that we haven't been tapping into.

The success of Roche's relationship with Genentech seems to have made it the model for pharma-biotech acquisitions, with the biotech allowed to maintain its day-to-day independence. Does AstraZeneca intend to use a similar 'light touch' with MedImmune?

"We want to integrate the discovery elements so that we tap into the best targets and the best science, because we are in the same fields, but we are going to run two separate parallel development pipelines.

"MedImmune can run a biologicals development pipeline – because they are good at that – they have all of the process development and manufacturing capabilities in MedImmune, and CAT had a lot of the early discovery capabilities. So, we have integrated MedImmune and CAT – and that is not light touch, they are integrated.

"If we can get it right, the scientists will interact and we will have joint target identification and joint therapy area strategies, but we will also have two parallel delivery arms. We will bring it together at the top with an R&D portfolio board who overview the total portfolio and decide where the money goes."

You must be one of the last surviving originals from the ICI days – how does that feel?

"It makes me feel old! I've been here 32 years, in a number of different posts. So yes, I suppose I must be one of the last in that group. But it feels like the company continues to grow and offer new and exciting challenges. And through all the changes, it is about discovering better medicines to make patients better. And that has never gone away.

"I was lucky. I walked into this company in 1975 and they said 'we've got this molecule called ICI 46 474', which eventually became tamoxifen, which I was asked to look after. I never intended to stay more than a couple of years, and imagined I'd go back to hospital medicine. But I stayed and it turned out to be the five most exciting years of my life.

"That caught my imagination and ever since then, I've had the luck to be in a company that has had the highest set of ethical standards, being led by people with enormous integrity, and with potential medicines in our programmes which were capable of changing people's lives," he says.

"It will never be a smooth process and we will have our ups and downs, but as long as we keep doing that, I believe this company will have a future."

Box: Ones to Watch in AstraZeneca's pipeline

Phase III

Zactima – solid tumours

Being studied for use against medullary thyroid cancer and advanced non-small cell lung cancer, with regulatory filing expected in the second half of 2008. Potential for use in broad range of cancers, and peak sales expected to exceed $1 billion.

Recentin – solid tumours

A second generation drug in the VEGF inhibitor class of Avastin, being studied in glioblastoma and renal cell carcinoma. Approval expected by 2010, with peak sales forecast to exceed $2 billion.

Motavizumab – virus infection

MedImmune's contribution to the late-stage pipeline, a monoclonal antibody for the prevention of respiratory syncytial virus infection in high-risk infants.

AZD6140 – anti-clotting

A second generation rival to Sanofi-Aventis' Plavix, with small head-to-head studies already conducted.

Dapagliflozin & Saxagliptin – diabetes

Two new diabetes drugs in-licensed from BMS and to be co-marketed. Saxagliptin to be filed in early 2008, and expected to exceed $1 billion in peak sales.

AZD4054 – hormone resistant prostate cancer

Potential first-in-class in area of unmet need with a good safety profile.

Phase II

Medi-528 – asthma

A monoclonal antibody from MedImmune targeting interleukin-9 (IL9) Delivered by intravenous infusion. Launch expected in 2010.

AZD6244 – melinoma

In-licensed from Array, the drug works by blocking MEK, a critical enzyme which regulates cell proliferation.

Potential to treat cancers of the lung, pancreatic, colon and thyroid as well as melanoma.

AZD 9056 – rheumatoid arthritis

First in a new class of ion channel blockers (P2X7 receptor agonist), the molecule is also being studied in inflammatory bowel disease.

AZD3480 – alzheimer's disease and schizophrenia

Being developed in collaboration with Targacept, the drug works at neuronal nicotinic receptor.

The compound is planned for development as a treatment for Alzheimer's disease and cognitive deficits in schizophrenia.

AZD0837 – thrombosis

PN400 – pain