In this talk, Peter Kim reflects on his experiences at MIT, Merck, and back in academia to examine the disconnect between academic and industry perspectives on drug discovery. He outlines three fundamental challenges of drug development—very low probability of success (well under 10%), long timelines (≈12 years), and high costs ($1.5–$2.5 billion per drug). Kim argues that academia’s emphasis on being first is misaligned with industry’s focus on being the best, noting that many blockbuster drugs (e.g., Lipitor) were not the first in their class. He proposes solutions: extending academic projects through startups, forming true partnerships that leverage world‑class academic expertise alongside industry’s medicinal chemistry, drug metabolism, and formulation capabilities, and structuring IP collaborations with co‑ownership while respecting legal inventorship rules. Kim also discusses licensing strategies, emphasizing the need to balance exclusivity for highly motivated inventors with safeguards such as narrowly scoped exclusivity and robust diligence clauses. Finally, he raises concerns about institutional conflicts of interest, especially in informatics‑based IP where inventors may be barred from using their own inventions in their labs, calling for greater transparency and policy work. The talk underscores the importance of aligning incentives, clear IP frameworks, and thoughtful conflict‑of‑interest management to replenish the drug discovery pipeline.
Chapters
Thanks, John. I thought that I might make a few comments from the different perspectives that I've had. As John said, I started out at MIT as a faculty member for many years and then moved to Merck and am now back in academia. And so I thought maybe I would reflect on my experience regarding the interactions between academia and industries. Since that's so crucial to the third part of the report. I, of course, will focus on drug discovery, and I'll just state up front that there are, of course, many other aspects of interactions between industry and academia, but my own experience is limited to drug discovery, so I'll focus there. And then maybe talk about some of my reflections on the interactions between inventors and the universities. Having been an inventor myself and having interacted with inventors who have challenges when they're dealing with the universities and what those issues might be that we should be thinking about in order to increase the pipeline. And then finally, if I have the Dean of Research and the President of the University here, to bring up an institutional conflict of interest issue that I think actually is worthy of further thought. As we move forward. So to begin with my experience regarding the interactions between industry and academia, and again I'll emphasize I'm going to focus on drug discovery. The first thing that I would start off by saying is that discovering and developing a drug is actually really, really hard. And I hadn't really appreciated how really, really hard it was until I actually went to Merck and saw how really hard it was. Just to put it in perspective, if you have a new mechanism of action and you've tested that mechanism completely in animals and you have a drug candidate which now survives all of the things that need to happen for not being toxic and actually being able to be applied to a human being, the probability that that molecule will become a drug is well lower, much less than 10%. So the probability of success of what we're trying to do is very, very low. And you talk about risk, there's a lot of risk that's inherent in the business. Second thing is that it takes a long time. The recent report out of the Tufts Group that compiles this data on a regular basis says that it takes 12 years to discover and develop a drug. And 12 years is a long period of time. It's a long period of time particularly when the probability of success is so low. All of which leads to the third point, and that is that it's very expensive. If you include the cost of failure as the Tufts group does, it today, the estimate is that it costs $2.5 $1.5 billion to discover and develop a drug. So all in, it's actually an investment of $2.5 billion to create a single drug. So we're dealing with something which is very low probability of success, very lengthy, and very expensive. And so the major observation that I would make with regard to the interactions between industry and academia is that there's a real disconnect in the perceived value of a discovery in academia. In academia, we of course want to be first. We want to be ahead of everybody else. We want to be the discoverer. We want to make sure we get a head start towards a program, and we think that when we do that, which is really, really hard, that that's really, really valuable. But on the other side, from the industry's point of view, the probability of success for what you've discovered, even though it's very, very ingenious, is actually very, very low. Given that it takes 12 years to get to market even if we are successful, the fact that you have a 1 or 2-year head start is not that big a deal. And the sort of chemistry that is addressed in most academic centers is really quite primitive compared to the medicinal chemistry and these huge departments that are expert at refining a drug to make— a drug molecule to make it into the characteristics that you want. So the initial leads or hits are usually not considered that valuable. And the final thing is that being first isn't all that important. What's really important is being best. Just to put that in perspective, Lipitor was the 5th drug to be in that class of categories, the 5th statin. So this multi, multi-billion-dollar success for Pfizer was actually not the first, it was not the second, it was actually the 5th molecule to come to market in that category. And so being first, the premium of being the discoverer, which is so important in academia, is much less important in industry. So what are the potential solutions to this disconnect? Well, one is, and I see actually more and more of this, and I think at Stanford more people are doing this, and it's really quite admirable, is to just carry the project further, either in academia or through a startup company, and take on that risk and actually increase the value. And I think that's perfectly appropriate. Provided that the academic understands that there's a high probability that they're going to fail as they move forward in this direction. But I thought I would bring up another possible solution which I really tried to push when I was at Merck, and that is to recognize that the value of a really good partnership is not so much in the initial discovery or hit, even though that's what the academic is really excited about, but rather it lies in the fact that you can work with the world's expert in a particular area. And that's something that you don't have. You don't have world experts in all the different areas in your company. And so the deep subject matter expertise of the professor who is the world's expert in X is actually something that is very valuable to a company that's trying to develop something. And the value of industry then is to bring the expertise in other areas that the world's expert in X would not be familiar with, such as medicinal chemistry, drug metabolism, animal toxicity, formulation science, other things that are often considered boring to the academic but are so critical to actually bringing the the product to bear. And I would just point out one thing that I think, if you can, is really important, and that is to make sure there's a true partnership with regard to IP. Intellectual property and inventorship, who actually gets listed as an inventor on intellectual property, is defined by law. That's not something that you can play around with. So you either have an inventor and the person satisfies the requirements for being an inventor or not. But ownership of IP is completely negotiable. And so you can have rules such that the ownership of IP is different than the rules about who is actually the inventor. And so the best sort of collaborations I felt were those in which you had a collaboration between academia and industry. And even though inventorship in the project would be defined by law, any IP that came out of that partnership would be co-owned. And so there was no question about who actually was going to benefit from either party coming up with an invention. And what that does, of course, is it opens up the doors, or at least it prevents them from being closed, so that people are not secretive with their partner, but instead are really trying to create value as we move forward. And so I think Certainly from the perspective of the industry, or certainly from the perspective of Merck when I was there, that was something that I tried to push, was actually having a true partnership with regard to ownership of IP. The second thing that I thought I would reflect on are the interactions between inventors and universities, and that's really what I would want to focus on there is exclusivity versus non-exclusivity. And I want to emphasize that I'm not talking about broad platform technologies. There are some inventions which we would all agree should be and absolutely must be licensed on a non-exclusive basis so that the whole world can benefit from them. But there are other inventions where it's not so clear, where we're dealing about specific enabling technologies or in the case of drug discovery with specific drugs or specific targets as we move forward. And understandably, the university feels, you know, as I say, completely understandably, that they want to maximize the value that they get out of that invention. And they also importantly want to make sure that the invention doesn't get buried or stuck because they've exclusively licensed it to one party that for whatever reason is not able to execute on developing that invention. The major point that I would want to raise for this audience is the importance of capitalizing on a highly motivated inventor, if one has a highly motivated inventor. Because a highly motivated inventor can actually really help move an invention forward and therefore bring not only capital in the form of money, but capital in the form of good deeds, or in the case of drugs, public health, through that motivation. And oftentimes the way to capitalize on a highly motivated inventor, and certainly from the investor's point of view, what they would prefer is an exclusive relationship. And so that then asks— begs the question, well, how do you structure things with an exclusive relationship when you have these other perfectly legitimate concerns. And I don't really have a good answer there. I think the Office of Technology Licensing does make an effort, and I think it's very important to try and carve out a specific area of exclusivity so that there are other options that are still available to the university for license. But the other thing which I see less use of, and maybe it's because it's just a lot more work, are diligence clauses that really have teeth and that are really specific so that if the licensor does not actually invest appropriately or if the investor does not hit certain milestones, that the university in an aggressive manner will come back and say you have not satisfied what it is that we need to have happen. So I throw that out as something to think about. And then the third thing, which I'll just very briefly mention because I've become aware of it recently, is now around the issue of institutional conflict of interest. And here I would expand the discussion to include broad platform technologies, and in particular intellectual property— excuse me, information technology intellectual property, informatics-based IP. I've become aware of several situations at Stanford where the inventor at Stanford starts a company and the technology or IP is available via a license from that company to everybody else in the world, including every other lab in the world except for the inventor's lab. And that's because of, I think, legitimate concerns about whether or not there's going to be a conflict of interest between the inventor continuing to work on that invention in his or her lab, for example, even using students, which would be highly inappropriate, I think we would all agree. And because of that concern though, we're in this funny situation where the inventor can't use their invention in their own research. Now some of my esteemed colleagues say tough, that's the way it is, and the person has to make a choice, but I would just throw out that that's something that perhaps one should be working on more in terms of trying to figure out is there a way around this. And clearly, I believe transparency is a critical piece of that in terms of figuring out how to make it happen. So those are just 3 areas I thought I would throw out to sort of give a perspective from my point of view. Good. Good.
