Can you fix a broken drug company research lab?
Jack Scannell, the European pharmaceuticals analyst at Sanford C. Bernstein, recently held a conference on the future of drug research and development. On the last day, he had representatives from two of the most successful drug development organizations on the planet: Sean Bohen, who heads early resaerch and development at Genentech, part of Roche, which has had a legendary string of cancer drug successes; and Mads Krogsgaard Thomsen, the chief scientific officer at Novo Nordisk, one of the dominant players in diabetes and the best-performing big pharma stock over the past decade. This story is based on a transcript of their talks.
Both executives raise a problem with fixing research and development that should scare executives at less productive companies: Sometimes trying to fix things can just make them worse. When you’re running a research lab, change is bad.
First, some stats from Novo to show how much success rates vary. Novo focuses mostly on protein drugs, like insulin. A little less than one in three of these medicines that starts clinical trials reaches the market. For other protein drugs not being made at Novo, including insulins and also anti-cancer antibodies, that number is more like one in five. For small molecule drugs, the lifeblood of drug companies like Pfizer and Merck, it is one in 25.
To what does Thomsen attribute his success? In part, it’s taking the long view, without much disruption to the researchers. He told Scannell’s conference:
[W]e don’t think in weeks or months. Maybe not even in terms of one, three, or five years. We think about decades. Personally, I have been heading up R&D in diabetes for 18 years. When I started as the R&D chief for diabetes in 1994, my dream was actually to create the perfect basal insulin, the perfect prandial insulin, and the world’s first really good GLP-1 based therapy (plus, of course, a little bit of other stuff). But the insulins and GLP-1 were top of the agenda. And actually, that is what today, 18 years later, we are on the verge of having delivered.
One result: the staff turnover at Novo R&D is just 3% a year; unwanted turnover is 1%; the wanted turnover (as in, Novo wants people to leave) is 2%.
Bohen, from Genentech, puts an even finer point on this: He says that R&D turnarounds are possible, but difficult, particularly because of the long timelines involved. And it’s possible, basically, to reorganize yourself to death.
An observation I will make about most of the industry is that they go through reorganisations so quickly that you can’t even figure out whether the last thing worked or not. It’s such a long business cycle. Some of these companies, I’m sure they had the right structure at some point in time and for the 18 months they kept it, it could have been very productive. But they changed it again before they ever found out. People just don’t have the patience to stick with it.
Scannell puts an even more negative tone on this. Perhaps, he says, companies with broken research and development divisions shouldn’t put money and effort into fixing them.
[R]eturns today on R&D at an industry level are almost certainly below the cost of capital, and appear to be on a downward trend. This means that an underperforming R&D organisation that labours for years to bring itself up to the industry average will still, at the end of the process, be wasting shareholders’ money. Even in this case, managerial success would equate with financial failure.
Whether or not — and how easily — drug research labs can be fixed when broken is perhaps the key question for investors in drug stocks. Sanofi is making a very public attempt to revive its labs. Will that work? Pfizer seems to be undergoing a research revival. Is it for real? AstraZeneca has had among the very worst research productivity in the industry, spending a staggering $11 billion in R&D money for each new drug approved. What should its next chief executive do to right the company?
It’s not true that research turnaround can’t work. Novartis certainly pulled one off a decade ago. More recently, Bristol-Myers refocused itself on cancer and other niche diseases, and is now seen as one of the industry’s most successful drug developers, according to a recent survey of buy-side analysts by ISI Group’s Mark Schoenebaum. But with research productivity dropping across the industry, laggards might want to think first before they try to disrupt themselves.