Here was the plan: give a microphone to one of the loudest voices about the drug industry’s inability to invent new medicines. And then let four of the top executives in pharmaceutical R&D respond.
The discussion went down at our first annual Forbes Healthcare Summit last week, and the critic, Bernard Munos of the Innothink Center for Biomedical Innovation, did not disappoint. Drug company R&D, instead of generating profits, returns only 70 cents on every dollar spent, he alleged. Companies are organized in such a way as to kill good ideas, not let them bloom, but at the same time they flood their pipelines with mediocre drug candidates in order to look busy, long before the development of those experimental medicines should have stopped.
“It's not FDA's fault,” Munos said. “It is management’s fault. Management responsibility for poor management decisions that are driving R&D cost.” (Videos of Munos’ talk, and the entire panel, in two parts, are embedded in this post.)
[newsincvid id="23915419"]
The battle cry that drug R&D is failing isn’t news – I wrote about it here and here – but it’s worth pausing to note the scale of the problem. The drug business is like the computer business in reverse. Since Gordon Moore coined his famous Moore’s Law in 1965, computer power has been roughly doubling every 18 months, leading to the personal computer, Google, and the iPhone. The drug industry’s equation is just as inexorable: the number of medicines invented per billion dollars in research money spent is halved every nine years -- an 80-fold increase in cost since the 1950s. Industry analyst Jack Scannell, writing in Nature Reviews Drug Discovery, gave this economic quicksand a name: Eroom’s Law. That’s Moore’s Law backward.
I knew that I was in for some interesting thoughts the moment Munos splashed Eroom’s law on the screen in front of 200-plus people and George Scangos, the chief executive of Biogen Idec, leaned over and said, “Over the past six years, it’s flat. It’s no longer getting worse.” And that was the view of the other panelists, including Paul Stoffels, who runs research at Johnson & Johnson, and Mikael Dolsten, who does so at Pfizer. In the words of Onyx Chief Executive Tony Coles, insisted that it is “an exciting and unprecedented time” and “an opportunity to redesign the way we do drug discovery and drug development.” The same day as the conference, the Food and Drug Administration announced that it approved 35 drugs in fiscal year 2012, the same number as the year before and a near-record take.
[newsincvid id="23914213"]
Is the drug industry really coming out of a fifty-year productivity slide and a ten-year slump that made it even worse? Here are some reasons for hope from my discussion with these research leaders, as well as some reasons I’m still worried that the worst is not over yet.
The Genomics Bust And Boom
One scapegoat for the research bust, according to this expert panel, was the attempt in the 1990s and early 2000s to use early data about the human genome to invent new drugs.
“In the '80s and the '90s, we did pharmacology research,” said J&J’s Stoffels. “In 2000, we had the genome, human genome. And we thought, everyone thought, that we were going to accelerate drug discovery by having all that information. But what we actually did is put pack from pharmacology to targets almost ten years. We created a gap. Now we have the know-how in new targets. And I’m convinced at this moment, there is going to be a significant increase of new products coming to market.”
[newsincvid id="23915420"]
Biogen's Scangos agreed: “The industry and the scientific community as a whole wasted a lot of time on the new genomics technologies when they were first available because they were exciting, they opened up new areas of study so people jumped on them. But they were immature, so there was time, there was money wasted. Now the technologies are way better than they were, people now understand how to use them, and we’re seeing the fruits of that. I’m convinced we’ve bottomed out and we’re on our way back.”
There’s even a good example, from Pfizer’s Dolsten, of a drug that has come directly out of these changes. Xalkori originally failed. But researchers discovered that it was incredibly potent in 5% of patients with non-small cell lung cancer, giving Pfizer one of its most important new drugs in years. There have been some other drugs that were clearly the result of genomics that have succeeded, including many other targeted cancer medicines and Vertex's cystic fibrosis drug, Kalydeco, but it's still early to say that these new tools have led to better medicines.
A New Ecosystem
What else might have been causing drug companies to produce fewer new medicines? Lots of people, in the industry and outside, blame size. Biogen's Scangos, who started his career at Bayer, said the problem is that at any R&D organization, a few key people are responsible for the majority of the progress. The bigger an organization is, the harder it is to identify those key people.
It's development, the conduct of huge clinical trials, where size and experience become an advantage. Stoffels remembered starting his own biotechnology companies focused on HIV in 1996 with $35,000 and five people. They were agile, fast, and independent, until the final year when he spent $80 million on studies in a single year. The second year after J&J bought his biotech's, it spent $400 million globally on two HIV drugs.
So why not just send all the basic research folks to work at smaller companies, and leave the drug giants to buy and develop them. One member of the audience even asked this of the panel -- and got a vehement response.
"I would say that's a bad idea," said Tony Coles of Onyx, which, while the smallest most nimble company on the stage is also the one that doesn't do basic research. "And it's a bad idea because I don't think we can leave these important questions of science to the random organization of a private company that will choose to focus on any number of diseases that could be big or small."
Instead, they all agreed, the new model is to have some really small research organizations but also big ones that can weather long journeys from idea to product. Dolsten pointed out that Pfizer's new rheumatoid arthritis pill, Xeljanz, took fifteen years to develop. In a phrase Dolsten turned after the panel ended, the ecosystem needs lots of different species. The problem is managing an entire ecosystem is far from easy.
Increasingly, Stoffels said, the model for inventing new medicines will look like what happened in HIV two decades ago, where drugs must prove themselves early in small tests (like killing a virus, or lowering a protein thought to be linked to disease), egged on by early cooperation on basic science and then competition to get the best molecule to market first. It's enticing idea. But one has to ask: if it's taken us eighteen years to learn those lessons, what else are we missing?
The signs on drug R&D are better now than they were before. But you could also say that in the late 1990s, before the industry went into a profound drought. There's reason for hope, but we know these problems are not cured yet.
The Forbes Healthcare Summit was supported by partners Medidata and Masimo and supporters MITRE, BASF, Verizon, Cancer Treatment Centers Of America, Arthrex, and Onyx Pharmaceuticals.
Video: Bernard Munos' Talk | Video: Pharma Research Heads Speak, Pt. 1 | Video: Pharma Research Heads Speak, Pt. 2 |
