...Modern drug discovery aims to amplify Fleming's serendipitous circumstances a millionfold: pharmaceutical companies search through combinations of molecular compounds at random, hoping to find a hit.
But it's not working as well as it used to. Despite dramatic advances over the past two centuries, in recent decades biotechnology hasn't met the expectations of investors -- or patients. Eroom's law -- that's Moore's law backwards -- observes that the number of new drugs approved per billion dollars spent on R&D has halved every nine years since 1950. Since information technology accelerated faster than ever during those same years, the big question for biotech today is whether it will ever see similar progress. Compare biotech startups to their counterparts in computer software.
[comparison table here]
Biotech startups are an extreme example of indefinite thinking. Researchers experiment with things that just might work instead of refining definite theories about how the body's systems operate. Biologists say they need to work this way because the underlying biology is hard. According to them, IT startups work because we created computers ourselves and designed them to reliably obey our commands. Biotech is difficult because we didn't design our bodies, and the more we learn about them, the more complex they turn out to be.
But today it's possible to wonder whether the genuine difficulty of biology has become an excuse for biotech startups' indefinite approach to business in general. Most of the people involved expect some things to work eventually, but few want to commit to a specific company with the level of intensity necessary for success. It starts with the professors who often become part-time consultants instead of full-time employees -- even for the biotech startups that begin from their own research. Then everyone else imitates the professors' indefinite attitude. It's easy for libertarians to claim that heavy regulation holds biotech back -- and it does -- but indefinite optimism may pose an even greater challenge for the future of biotech.My readership is far more knowledgeable about pharma than I am. Suffice it to say that Thiel's characterization of modern medicinal chemistry is a lot closer to, say, Marcia Angell's understanding than it is to someone in the industry. "combinations of molecular compounds at random" Eyeroll.
What I am really stunned by is Thiel's suggestion that biotech startups having insufficient willpower and intensity for success. Most of the time, they're structured just like software startups, with a staff of smart, hungry young scientists who are fully employed by the companies and strongly incentivized by their pay structure to make big impacts and meet their milestones. I'm not really sure that it matters that the professors are only consultants -- most of the time, it's the relevant laboratory folks (postdocs, grad students) that get hired in full-time to really get the laboratory work started.
Why does Peter Thiel think that software startups have sufficient willpower to wield the Green Lantern ring, but biotech startups do not? I'd love to know what kind of interactions with biotech that he has had that he's come away with such a poor (and misinformed, in my opinion) viewpoint. We'll never know, I suppose.
It reminded me instantly of the hilarious "Dabblers And Blowhards": http://idlewords.com/2005/04/dabblers_and_blowhards.htm
ReplyDeleteHalcyon molecular I believe. Not really biotech, but probably close enough in his eyes.
ReplyDeleteI actually spoke with Halcyon CSO 4 years ago - he called me up to find out if I would like to work with them. My impression was that they had plenty of willpower and optimism but the obtuse Nature did not share their vision...
ReplyDeleteYeah, his analysis misses the whole "time to market" issue. If software or apps had a 12 year approval cycle I bet it wouldn't get quite the IPO success rate it currently enjoys. Lack of willpower? My proverbial back end.
ReplyDeleteThe author reminds me of the parable of the blind men describing an elephant... (http://www.jainworld.com/literature/story25.htm)
ReplyDeleteThe next person to mention Moore's Law in reference to drug discovery...
ReplyDeleteMoore's Law in reference to drug discovery.
DeleteNow what C&EN Onion?
I wonder if there's as much incentive in biotech/small medchem startups as with computers/software. With biotech/med chem, you need so much money to come up with anything that initial investors are likely to have their investments diluted (though that would depend on structuring, I assume, about which I know nothing). In addition, the ability for employees to cash out their equity is pretty limited (these comments - http://pipeline.corante.com/archives/2012/03/01/what_sanofi_thinks_about_you.php#815142, http://pipeline.corante.com/archives/2012/03/01/what_sanofi_thinks_about_you.php#815350, and http://pipeline.corante.com/archives/2012/03/01/what_sanofi_thinks_about_you.php#815832) - in addition, it seems like (but don't have quantitative data, so time for a salt lick?) that small pharma employees get laid off if the company fails, and laid off if it succeeds (either to spend money on develepment of candidates or because they were bought and the bigger company only wants their pipeline and IP). On the other hand, since the financial engineers funding both are the same, it could happen in computers, but there they are likely to have less leverage because of reduced need for funding.
ReplyDeleteI think Moore's Law wouldn't be happening to computers if they had to be tested as much as drugs, and had to be as faultless. That little bug in your new app would be harder to swallow if instead of revealing your computer's innards, it blinded you or made you nauseated or killed your liver. And, of course, as Milkshake said, Nature and living things don't care about willpower - they work, and do not reveal their secrets because lots of people want them to.
I've seen a lot of these sorts of comments by outsiders about how industry just randomly looks through a bunch of compounds hoping to stumble onto new drugs. I wonder if part of that is because high-throughput screening was a flashy sort of thing that people liked to show off (high-level presentations highlight the new state-of-the-art multimillion dollar robotic facilities, and the tour groups don't stop at the old, dusty, and cluttered chemistry labs), so that's all that these people took away from drug discovery. They don't realize that it's only a *starting point* for evaluating a hypothesis that people have already put a lot effort into, and that once the hits are found, most of the work happens after that. I don't know how to address this, except trying to educate people about what medicinal chemistry actually is. Better communication, tell the success stories, and tell the failure stories, which are probably even more important to convey since it would help show that we really don't know the answers so we keep trying things and learning from them.
ReplyDeletePeter Thiel's ignorance is frustrating, but I don't know that "combinations of molecular compounds at random" is so far from the truth. Certainly, to my mind, it was true in the combi chem era which is not so distant. I'd even suggest fragment based DD can incorporate an element of randomness to it. Until we can predict why ethyl groups can be better than a methyl or isopropyl group (for example in ICPT's OCA derivative) it seems to me this is a reality.
ReplyDeletePaypal is, I'm sure, a fine creation but I very much think it pales in comparison to insulin or Solvadi or other drugs that really help people, rather than make it easier for them to bid on Beanie Babies.
I think my concern is (similar to "z" above) that Thiel seems to think that drug discovery begins and ends at HTS.
Delete