I encountered John LaMattina’s pharma-blog, Drug Truths (johnlamattina.wordpress.com/), this summer after he penned an article in Nature Reviews Drug Discovery in which he discussed the chilling effects that large pharma mergers have on R&D. Dr. LaMattina wasn’t some random pharma-griper; he served as senior vice president, Pfizer and president, Pfizer Global R&D, the capstone of a 30-year career at the company, where he began as a medicinal chemist.
Since leaving Pfizer in 2007, Dr. LaMattina has penned a book, Drug Truths: Dispelling the Myths About Pharma R&D (www.amazon.com/dp/0470393181), and has taken on a senior partner role at PureTech Ventures. I enjoyed his writing and asked him to sit down for a conversation about the state of pharma R&D and more.
Contract Pharma: You’ve been in the industry for a long time, so what’s your perspective on where pharma R&D is going?
John LaMattina: Contrary to popular opinion, I think it’s a great time to be in pharma R&D, both in large and smaller companies. The amount of knowledge that exists right now, in terms of understanding of diseases, is light years ahead of where it was when I started. The technology one can drawn upon for rapid generation of understanding activity of molecules, both from the chemistry and biology standpoints, is also light years ahead from when I entered in the mid-’70’s. So I’m envious of the young people who are picking this field for their careers.
There are some greater hurdles now than existed back then, too, in two ways. First, FDA has stricter requirements for approval, especially when it comes to long-term safety data. This would’ve been unheard of 10 or 12 years ago. Any drug in development for diabetes has to have not just short-term plasma glucose-lowering data, but also long-term outcomes. The FDA’s saying, “So what if you can lower plasma glucose? There are plenty of drugs that can do that. What are the long-term effects?” They want to know if this lowering actually reduces the risk of heart attacks and strokes.
We see the same thing in cancer. Just because you’re shrinking tumors doesn’t mean you’re saving lives. FDA now requires long-term outcomes. If it only increases a patient’s life by a month, maybe it’s not worth approving. If it increases survival by a year, that’s another story. In the old days, it was assumed that if you were shrinking tumors, a long-term benefit would accrue.
CP: So you’re saying that we were essentially following inaccurate biomarkers?
JL: Surrogate markers are indicative that something may happen, but they don’t always lead to the outcome you hoped for. As a result, those markers are helpful in guiding development, but not very beneficial in getting drugs approved. Unless there’s nothing else to treat a certain disease and the agency wants to get something out there to try to help patients in dire need.
CP: You mentioned two hurdles for modern R&D. What’s the second one?
JL: The second hurdle is commercial potential. Now we have to deal with a host of payers. In some countries, the government is the payer. In the past, you could launch a new drug that was similar to existing ones, and patients might try it. Now payers are saying, “We don’t know if your new and expensive drug is much better than what we’re already prescribing.” You have to really build into your long-term clinical programs not just safety and efficacy studies, but also outcomes studies. You have to compare not just to placebo or a decades-old treatment, but to the standard of treatment. Then you can justify that the drug is worthwhile and get it added to formularies.
Those are the two major changes that have occurred and are driving up R&D costs. They lengthen the period of time it takes to develop a new drug and cause a lot more late-stage attrition. In the old days, about 90% of drugs in Phase III would reach the market, because you knew so much about the drug by the time it reached that stage. Now your Phase III program will have differentiation aspects built in, as well as three-year outcomes studies. At the end of that period, after all that expense, you might look at it and say, “It’s just not worth launching this.”
CP: Do you think we’re exiting the blockbuster era for pharma?
JL: You can’t develop a compound that has a potential of $200 or $300 million in annual sales if you’re selling to a general practitioner market. The outlay of money in terms of advertising and office visits means it doesn’t make financial sense. But take Xalkori, Pfizer’s new lung cancer drug. It only works in around 5 to 10% of people with lung cancer, but when it works, it works very well. Thanks to the accompanying diagnostic test, oncologists now have a drug that they can prescribe that they know will work with this type of patient, and payers know that they’re not paying for something that’s hit or miss.
For years, people would ask if “personalized medicine” is the death of the industry. If you had a drug that payers, patients and physicians had confidence would work, and didn’t need broad marketing and detailing, you’d be able to price it very well. Drugs like Xalkori can form a new category called “niche blockbusters,” as coined by Jonathan Rockoff in The Wall Street Journal.
The ROI for these drugs would be very good, because you could price them effectively. Now, we’re not going to see a $13 billion a year drug like Lipitor again; that was an aberration. But you’re going to see more multi-billion-dollar drugs, depending on the indication. I think it’s potentially a very good time for the industry.
CP: What tension exists between R&D and commercial operations, when it comes to conveying this idea that you’re not necessarily going to hit the home runs that we saw in the previous two decades?
JL: I think it comes down to the CEO and his board. Boards tend to be very susceptible to criticism from analysts. When analysts talk about how ROI for R&D is abysmal, so R&D spending should be cut, it sends up alarms. You need to analyze why R&D return hasn’t been as good, why it was great in the ‘90’s, and also why it was abysmal in the ‘80’s. This is a cyclical pattern for whatever reason, but it’s the engine of pharma, and they need to invest in it.
Those CEOs that have boards that agree with this line of thinking are in better shape than those who don’t.
CP: Who are your heroes, in that respect?
JL: There’s one name that jump to mind: John Lechleiter at Lilly. He’s a chemist like me, so I’m partial. He’s been pretty adamant that cutting R&D is nuts. Wall Street has punished him for it, when he’s investing 18% of the top line in R&D. He’s saying, “I want Lilly to innovate its way out of its problems.”
I hope guys like John will win. Boards put pressure on management, but boards understand the business side more than the R&D side. That’s particularly prominent in U.S.; some European companies — like Roche and Novartis — have a more enlightened view about the importance of R&D.
CP: How difficult is it to run R&D at a large company? It seems that it’s easier to say no than it is to justify putting something innovative into the clinic with the risk of failure, and that the larger the institution, the more likely for that risk-averse mentality to prevail.
JL: Scale has tremendous advantages that you have to capitalize on. At Pfizer, we didn’t necessarily have armies working on our research. I was talking about this once with Colin Goddard, the chief executive of OSI, which we worked with in the 1980’s and ‘90’s. The biggest R&D group we ever had at Pfizer at any one time was about 220 scientists, chemists, biologists and others, and that was in oncology, by far our largest group. They made some pretty good discoveries. Colin thought that was too small. At its peak, OSI had more than 400 people working in oncology, he told me. It shows you that a smaller company can have almost double the people working in a therapeutic area as Pfizer.
So size can be an advantage. The bigger aspect of bureaucracy is the phenomenon of people thinking they can predict what will be a winner and what will fail. I can tell you horror stories of how drugs that were projected to be major blockbusters turned out not to amount to much —
CP: — Like Exubera?
JL: Oh, that’s a classic example. On the other side, when we did the deal with Warner-Lambert for Lipitor, our people predicted peak sales to be around $800 million a year. We never dreamed it would reach 15 times that number. Stars aligned around that; not only was it a great compound, but we also ran great post-approval studied in different groups. It changed medicine in a lot of ways.
With Chantix, our smoking-cessation compound, the commercial guys telling us it would never be bigger than $100 or $200 million a year. The only reason we went forward was it what that we believed it would be a good thing to do. It’s important to help people quit smoking. And it grew to a solid $700 million a year drug.
CP: To what extent were you involved in outsourcing decisions at Pfizer R&D?
JL: Well, there was a big R&D budget at Pfizer, but the actual part that I controlled was lot smaller than people thought. Commercial division controlled a big portion for post-approval studies. Still, we’d try to stretch that budget as far as we could. On one side, we would outsource functions like IT, knowing that we could make dollars stretch a lot further, and devote ore money to trials.
When we did all those mergers, we inherited a lot of toxicology facilities. Initially, we were doing all our tox studies in-house. We got to a point where we realized we could keep half of those facilities, and use CROs to handle more of the routine studies.
Outsourcing has its place, but you can’t outsource everything. It certainly has a role and it makes a lot of sense in particular areas.
I would really try to avoid building captive facilities. I wasn’t in favor of putting up a research center in China. I said, “That’s good for right now, but eventually this market is going to command labor costs as expensive as those in the U.S., once standards of living rise.” Maybe the next market was going to be, say, Vietnam. So I tried to keep as much flexibility as I could. We’d like to outsource to those areas, but keep the flexibility to move to other areas.
CP: How well do you think large pharma is handling the de facto outsourcing of drug discovery, in terms of in-licensing or acquiring compounds from emerging and small biopharmas?
JL: The thing to understand is, it’s not new. All of these big companies, for decades, had about a third of their revenues generated by compounds that were in-licensed. From the 1970’s on, these came from smaller pharma companies, or from Japanese or European companies that needed a partner in the U.S. Those companies have disappeared over the years due to mergers, failures, etc., but they’ve been replaced in the biopharma ecosystem by startups.
No matter how smart you are and how much R&D funding you have, you can never corner the market in good ideas. You’ll come up dry sometimes, so you’d be foolish not to be on the lookout for external compounds to complement your internal pipeline.
It’s starting to extend beyond biotech startups and into universities, as they try to capitalize on some successes. There’s a tremendous pool of ideas to tap into, and I still subscribe to the “rule” that you should in-source around a third of your pipeline. Companies need to do this aggressively, both in early and late stages.
CP: Pfizer’s greatest success comes from an external compound.
JL: One of the greatest stories of of external pipeline development is Lipitor. It was being developed by Warner-Lambert while I was at Pfizer. One of my best friends from grad school was head of chemistry at WL at the time. He and I were at a meeting and he said, “We’re struggling with whether to put our statin into development.” At best, it would be the fifth on the market. I told him, “It’s a pretty big market.” I thought, even if it made $100 million, that would be pretty significant, so that was a good reason to go into to development. Then, when it got into the clinic, it turned out to be way better than everything else in its class. There was no reason to think that would happen, based on the animal data.
CP: Is there an optimal range of R&D spend against top-line revenues?
JL: A lot of it depends on the stage of a company’s key products, what it’s trying to accomplish, etc. Lilly’s on the upper end at 18 or 19%, which may be high, but 9 or 10% is a bit low. It’s interesting to note that Amgen, one of the few successes in the biotech arena to become a fully integrated company in its own right, has announced R&D cuts recently. Analysts came out with “Well, we always felt that $3 billion was too high for R&D, and it’s around 19% of revenues.”
CP: They didn’t feel that way until sales began to decline for Aranesp and Epogen.
JL: And that’s the gist of what I’m saying. What right does Wall Street have to dictate what R&D spending should be? They want to see short-term boosts. I don’t think Wall Street should have that much influence on research investments.
CP: When it comes to those boards and CEOs, have any of them commented about your books and blog-posts?
JL: As a matter of fact, no! I received a lot of praise when I wrote an article for Nature about how mergers affect R&D. I got a lot of e-mail from people at different companies (or former staffers who had been laid off after mergers). One had the subject line of “Bless you!”, but nothing from anyone in a position of real power.
It’s not going to change my views.
CP: Tell me about your new role at PureTech Ventures.
JL: I am a Senior Partner. Two things attracted me to the company. First is its business model. It’s not a venture capital investor. I talked to several of those, but PureTech is unique. It doesn’t wait for people to come to it with ideas; Pure Tech starts companies. They target an indication, say, schizophrenia, and say, “There has to be a better way to approach this.” They’ll bring in leading academics in that area and ask about the drawbacks in the current model of treatments, and the latest advances. After a few of those sessions, PureTech will try to license the IP for good ideas that come out of them, and hire some staffers — like former colleagues of mine at Pfizer — to run these virtual companies. The goal is to get to a certain decision point, using minimal resources, and then try to take the company to a VC or a pharma partner.
The other thing that drew me to PureTech was the other senior partners. There’s Bob Langer, who started something like 20 companies over the years, Ben Shapiro, who used to run discovery research at Merck, John Zabriske, who was the CEO of Pharmacia Upjohn, and a few financial people. We have a monthly call and quarterly meetings in which we’re bandying about ideas with some of the best brains in the industry.
It’s like being part of a discovery group again.
CP: Any good news come out of it so far?
JL: We have some exciting new approaches. I’m involved with a novel obesity treatment through a company we started called Gelesis.
The best treatment for obesity is bariatric surgery, but you can’t do that on millions of people. So we’re developing polydextrose compounds that mimic the effects of that surgery, rather than going after metabolic pathways. You take some capsules before a meal. When they get to your stomach, they absorb water and swell up, giving a sensation of being filled. You can get enough nutrients from a regular meal, but you won’t overeat. Reducing food intake reduces obesity.
The beauty of the material is that its water-retaining properties are pH sensitive. As they move down the GI tract, the pH changes and the water gets released. It’s a terrific approach. We’re talking to various companies that are interested in licensing it. That’s one program that I particularly love.
CP: How different is it to work on this scale than your days of multi-billion-dollar R&D budgets at Pfizer?
JL: I first got into this area because I’m a scientist. I’m probably doing more science now than I was in my last few years at Pfizer, where I spent a lot of time dealing with site closures and downsizing. These companies really value my input and experience from large pharma. One of the great things about directing R&D at a company like Pfizer is that I got to keep track of hundreds of projects. When you get that experience, you learn what works and what doesn’t, what pitfalls may be out there, and ways of approaching problems.
CP: Where do you think some of the biggest R&D opportunities are?
JL: I believe cancer will be a chronic disease within a decade. There's something like a thousand compounds in the clinic right now for cancer treatments, along with the drugs that have been approved in recent years.
I envision a patient getting diagnosed with cancer, and his oncologist taking a biopsy of the tumor to get an idea of what the main driver for the tumor growth is, then being able to prescribe a specific, targeted drug that will deal with the aberration in your cells that's causing the tumor. You'd probably combine that with something to prevent tumors from growing blood vessels, as well as something to up-regulate your immune system. It's similar to what we learned about triple-therapy to treat AIDS.
This will occur with in a decade. We might not have a cure, but we'll have a treatment.
CP: Were you around for torcetrapib?
JL: Boy, was I. My life was all-torcetrapib, all the time, in those days. We started that program in 1990, when I was in charge of chemistry for the metabolic disease area, which included atherosclerosis. It was a very difficult project to get off the ground. When we finally discovered torcetrapib, the project really took off. We couldn't have an analyst meeting without my getting asked about its progress.
I don't think the field is dead. Merck and Roche are taking their compounds along, and we'll likely know in a year or so if the hypothesis holds. The fact that they've gone so far is a good sign, I think. If a lack of improvement in outcomes was there, it would probably have shown up by now. If elevating HDL works, they should see fewer heart attacks, strokes and procedures in the patient group. I have high hopes that Merck's study will work.
CP: So you think it was the particular compound, not the mechanism of action?
JL: We were dosing people for a long time, so we knew there were blood pressure increases. We never dosed it optimally. We dosed it in ways that gave 60, 70% increases in HDL, but as people were on it for a long time, they'd get a 2 or 3 mm increase in blood pressure. We learned that when people were being dosed for a few years. We had to stop the study when we learned that there was an excess of deaths in the torcetrapib group vs. the control group. I think it was due to the CV issues the compound had: it caused BP elevation and we couldn't give the optimal dose to get what Merck's getting, 100, 120% increases in HDL.
It's an expensive experiment: $400 million or so to learn that lesson.
CP: Occasionally, we get nutty MBAs declaring that pharma is only good at development and marketing, and thus should cut all early research and in-license everything. What’s your take on these more radical notions of pharma R&D?
JL: I just wrote about it recently on my blog, in relation to pressure from shareholders and boardrooms. I believe you have to have a core strong internal scientific group for two reasons. First, you have to be able to analyze what’s going on with a compound. Second, it becomes an expensive business model. If you’re licensing everything in, you’ll be paying a ton for these drugs, in both upfront costs and in royalties. Plus, your risk won’t be reduced. Studies have shown that startup companies have no higher success rate than big pharma, and might actually be worse. So you’d have more risk and more failure and possibly more cost.
Something to keep in mind is that startup companies have limited resources. They’ve got to do as much work as they can with much less, compared to big pharma. When we would review an early-stage program, we had a key phrase to keep in mind: “There are no show-stoppers here.”
Meaning everything so far is okay, but if we had had it internally, we’d have done a half-dozen more tests by now. So we don’t have answers to those other tests. There would be other bidders out there. We could ask them to run those other tests and get back to us with the results in, say, six or eight months. But they could just go down the road to another big pharma that might be willing to overlook the risk and buy the product outright.
That’s the kind of bind you get into, on the in-licensing front. It drives up risk.
CP: I suppose it highlights the tension between VC-funded startups, which are always looking at an exit strategy, and publicly held large pharma, which are driven by quarter-to-quarter statements and shareholder expectations. To that end, do you think we’re ever going to see a biotech make it to IPO again? Are they more likely to sell out to a large pharma than to hold onto a product and go public?
JL: I’m on the board at Human Genome Sciences, and we just had a very important product approved for lupus, with great potential in other immunological areas. We hope that’ll buy us some time to develop other products, but it is a struggle.
Someone told me that many startups manage to get one product to the market, but their second products tend to be something that they brought in. It’s a great strategy, but it shows you how difficult how it is to develop an internal pipeline.
Gil Roth has been the editor of Contract Pharma since its debut in 1999. He can be reached at firstname.lastname@example.org.