Published online 16 February 2010 | Nature | doi:10.1038/news.2010.74


Drug discovery for the masses

To sustain innovation, pharmaceutical companies will have to change the way they do research, says Derek Lowe. But does anyone know what changes to make?

Derek LoweDerek Lowe.T. Hashemi

The pharmaceutical industry is in a strange position: it continues to make tremendous profits, but many insiders seems convinced that its business model is unsustainable. How can both of these beliefs be true? The industry is built on assets that lose value over time — namely its patented drugs. And these take many years to make their way to the market through the research and development (R&D) pipeline — if they emerge at all. So things can seem fine, but trouble is always looming on the horizon.

Looking at the statistics for new-drug approvals, it's clear that the industry is not getting any better at bringing compounds to the market. This has many people searching for a better R&D model, and one proposal that keeps coming up is for some sort of 'open innovation' system. Not everyone means exactly the same thing by that term, but the general idea is that companies would allow others to contribute to their research efforts, sharing the risk in exchange for a share of the profits.

To some extent, that's already happening. Large drug companies are increasingly making deals with smaller ones, and will partner with their peers when the risks and returns look right. But mergers, buyouts and partnerships aren't the only things that open-innovation advocates have in mind. There's talk of more fundamental changes in how projects are advanced and how technologies are deployed, including creating partnerships with university labs or private foundations and assembling consortia to deal with large data sets that individual companies might not be able to handle. There's also the idea of 'crowdsourcing' — outsourcing proposals to large groups or communities to let anyone with an idea have a go at cracking the problem.

But in order to help, such techniques will have to be deployed against the rate-limiting steps in drug development, which vary widely. Some therapeutic areas have a shortage of good drug targets, whereas others have plenty of ideas but few that yield any molecules worth developing. In drug-discovery areas better-served by targets such as diabetes and cardiovascular, safety concerns are paramount, whereas in oncology, drug efficacy rules.

Changing drug culture

Still, there are some universal themes for drug discovery. In the early stages of the pipeline, many agree that we could use approaches that bind molecules to protein surfaces rather than to pockets adapated for small molecules, as has been traditional. And any approaches at all to the more esoteric drug combinations — such as those targeting protein–nucleic acid complexes — would be welcome.

There is a lack of fundamental biological understanding about almost all of the diseases that the industry would like to target — even things that might seem to be well understood. (For example, if you know a good way to safely raise the level of cholesterol-carrying high-density lipoprotein in patients with cardiovascular problems, the industry would love to hear from you.) And towards the end of the pipeline, human toxicology seems to be the same terrifying game of chance it has always been.

So ideas that address large-scale biology are probably worth trying, although one might wish first for a way to comb insights out of massive data sets. The industry keeps hoping that its problems can be solved if it can just collect enough information — some can be unravelled in this way, but it's not a straightforward process. Perhaps that's where the crowdsourcing theme comes in. No one knows where good ideas might come from, so why not invite them from everywhere?


That brings up another fundamental opportunity — or fundamental problem, depending on your point of view. Scientists who have worked in both small and large drug-research operations frequently contrast the two, especially regarding how new ideas are received. If some of the new consortia turn out to be huge organizations themselves, will they pick up some of the same bad habits? There are very few good insights that can't be killed off by interlocking layers of review committees. This makes one wonder about the fate of crowdsourced proposals. Getting ideas is a big first step, but recognizing them and following them up are even bigger ones, and some companies seem to have trouble doing that even with their own internal proposals.

But transforming ideas into useful compounds is no easy task. Proposals for open innovation often quickly call for the drug companies to change their research cultures. But the industry's chief executives have been saying the same thing for years now, to little effect. Some of the cultural changes seem, if anything, to have been for the worse. How the large companies are supposed to make the transition is never explained.

Perhaps the prospect of all those patent expirations, drug failures and price controls will concentrate our minds wonderfully. Opening up the labs is a strategy that goes against many strong industrial instincts. Seeing these ideas discussed seriously is proof enough that the pharmaceutical companies are facing some serious problems. 

Derek Lowe works on drug discovery in the pharmaceutical industry. You can read more on his blog, In the Pipeline.

Commenting is now closed.