Q&A with Avak Kahvejian on the Future of Biotech

Liz Pavone: Hello, I'm Liz Pavone, manager of content and science communications at Flagship. I'm here with Flagship General Partner, Avak Kahvejian, a succinct title for someone making a broad impact as an inventor, entrepreneur, and CEO. Since 2011, Avak has been a partner at Flagship where he has co-founded more than ten companies, including Seres, Ring, Cellarity, Generate, ProFound, and Ampersand. Welcome, Avak.

Avak Kahvejian: Thank you, thanks for having me.

Liz Pavone: We are recording in advance of the 2024 J.P. Morgan Healthcare Conference in January. Looking back on 2023, what were some of the big milestones for you and your team?

Avak Kahvejian: Well, one of the most recent milestones in 2023 was regarding ProFound Therapeutics, one of the companies we launched based on the appreciation that the human proteome was much larger than previously believed. There we hired a CEO and a CSO in September and we brought on John Lepore from GSK. He had 17 years experience there, where most recently he was leading all of R&D. And, as CSO, we brought on Klaus Urbahns a multidecade veteran of the drug discovery space, chemist by training, who most recently was at Merck KGaA, leading their research and development. So, that's a big catalyst for a company like ProFound because it essentially moves us into the next stage of expanding our drug discovery and drug development activities, building on top of the work we did on the platform, and moves us into the next year for business development, fundraising, and other activities. And then we launched another company, Ampersand Biomedicines, which is pioneering the concept of being able to program biologics, to program them to go and act where needed and not to act where they're not needed. And that was a really successful company launch. We also hired a CEO for that company. Jason Gardner joined us where prior to that he was CEO of Magenta, a startup, and before that had a long tenure at GSK. So, these are two major developments from my Pioneering Business Unit, two New Cos, that are off to a great start.

Liz Pavone: Programmable medicines is becoming a familiar term, at least here at Flagship, why is that? And what is the difference between a programmable medicine and a traditional therapeutic?

Avak Kahvejian: Well, traditionally, therapeutics have been developed or created in a very empirical way with a lot of trial and error. We've used a numbers game and a game of attrition to find the drugs that might work and then move those along to the clinic. And that has been obviously fraught with a lot of failure by definition almost. And it's been due to, mostly a limited understanding of how biology works and also the modalities at our disposal, the types of molecules we were working with, which were small molecules found in nature or antibodies generated in some animal model. The future now, which is the present for us, is the idea of having modalities that can be encoded, that can be programmed, as we're stating, to achieve a deterministic outcome or an outcome that's predictable. And, nucleic acid medicines have played a big role there. And that's been the first foray, if you will, in creating programmable medicines. But that notion of programmability is now moving into the realm of biologics, moving into the realm of small molecules. And what I mean by that is encoding the pieces of the biologic or the small molecule, designing it with a purpose in mind, and not trying a billion of them or a million of them to see which one sticks, but actually creating one with more of an engineering mindset and a predictable mindset.

Liz Pavone: Rational design is how we build bridges, planes, and my smartphone. Why weren't we developing medicines like this before?

Avak Kahvejian: Well, as I alluded to, I think one part of it was our understanding of biology. Those other things that you mentioned are all man made. And so, therefore, they're amenable to basic principles, and they rely on the rules of physics, which we started to harness and understand almost completely at the turn of the century, at the turn of the prior century. So in this century, I think we're seeing a decoding of biology such that we can start to apply those rules that govern the biological processes, that govern health, that govern disease, and then trying to impinge upon them, reverse them, alter them, and then using the modalities that we have at our disposal. The armamentarium is getting better and better, or broader and more diverse, such that we can start to design the drugs to impinge upon those programs. So, understanding the biology, being able to read biology, being able to write biology, has been a huge change in how we approach drug discovery and development.

Liz Pavone: What you had described were programs that we're impinging on, those programs that I suppose our bodies are running on, correct?

Avak Kahvejian: Right. So, our development is a program. Going from an embryo to an adult is a program. Our immune system runs on a program. It sees an antigen, it responds, it activates a whole cascade of events that are more or less, predictable, repeated, and by understanding those, we now know what the triggers are and what the outputs are of these processes. But this is only the beginning. There are so many other programs that are going on. The human body wasn't designed by man, it has evolved. And so the programs are much more complex than one can imagine, or maybe capture on a map, on a wall, on a wiring diagram. We've tried that with biochemistry before. It works. But when you get into more and more complex gene networks and cellular networks and physiology, there's a lot of interdependencies, redundancies, and we can't keep that all in our heads or just maybe map that out on one two-dimensional picture. So, with the advent of AI, we're now able to assimilate a lot more of these data points and the dynamics of how these things change and extract and elucidate programs. Then we now know, arguably, how to trigger them, how to stop them, how to slow them down, for example.

Liz Pavone: So is that the interconnection between the programs we're running on and these programmable medicines?

Avak Kahvejian: That's exactly right. But you need to have the ability to impinge upon them, as you said. So understanding them is one thing, but then bringing in something to alter them, or trigger them, or stop them is another story altogether. And doing that, with a high degree of efficiency and safety at the site where you want it to happen, and not elsewhere, is the other challenge. And that's where that toolkit of different kinds of molecules that we're working with comes into play. And the place where it was easiest, arguably, to program was in the realm of nucleic acids, because there we know the codon table, we know which triplets to arrange to generate a particular protein inside the human body. And the other biological program that's probably well-known or well understood is the immune system, as we said. So presenting an antigen to the immune system, we know will launch a cascade. And so that program, albeit, I guess simple to articulate here, but a complex one, was easy to trigger with a nucleic acid. And then the rest of the biological programs in the body did the job of making the protein and then reacting to the protein. So that's one program or set of programs, but as we move into other disease areas and other types of mechanisms we want to engender, we can't always use the same tool for the same types of programs. And so things that happen outside of the cell, for example, you may want to use a biologic for things that happen inside a cell, but have nothing to do with creating a protein, or removing a protein, you may need to do with a small molecule, and so having the ability to program those molecules to go where they need to go, to engage with the targets they need to engage with, and to know which targets you should be engaging with — all of this requires knowledge, understanding, and the ability to engineer the molecules appropriately.

Liz Pavone: And so it sounds as if we're in a state of transition to a new paradigm, a future space. What does biotech need to do to get to that next leap forward?

Avak Kahvejian: Well, we are looking at it at Flagship from a variety of perspectives, building bioplatforms that can achieve both the understanding of biology, and bioplatforms that can create the modalities to impinge upon biology, and then looking at also the intersection of those, whether that's through collaborations or whether some of those things are housed within the same company. But I think that's where the biotech industry as a whole needs to move, and to apply computational tools as much as possible, because the amount of information and the complexity of the information and the understanding realm is immense. And then on the modality side, optimizing the modalities, knowing why one modality works better than the other, or one design works better than the other, instead of leaving that to chance and trial and error, using computation to extract the best features that optimize multiple parameters in terms of pharmacology, efficacy, safety, stability, so on and so forth, requires also a computational mindset, and one that's taking in a lot of data, a lot of multiparameter, multimodal data, to optimize your modality. So biology has to move into that realm of becoming much more of an information science, and one that takes advantage of, what machine learning can do for us, which is give us superhuman capability to extract what's important from these data and to act accordingly. So that's where we're going at Flagship but I think the industry as a whole is moving into this more convergent mindset.

Liz Pavone: So, Avak, what are your objectives for this year's J.P. Morgan conference?

Avak Kahvejian: Well, as always, we're excited to get together with the rest of the biotech community and to showcase some of the cool bioplatforms that are emerging from our work here at Flagship. I think we want to use the opportunity to engage with pharma in particular, to essentially exchange ideas around these bioplatforms and the breadth of possibilities that can emerge from these technologies and to see whether we can start the seed or plant the seed for eventual partnerships with pharma. And this applies to any of the companies. that we're working with, whether that be ProFound or Ampersand or others.

Liz Pavone: So to conclude, I wanted to go through a lightning round of questions. Favorite Flagship-ism?

Avak Kahvejian: I like paranoid optimism.

Liz Pavone: Moment you knew you wanted to be a scientist?

Avak Kahvejian: As early as I was maybe four or five.

Liz Pavone: Moment you knew you wanted to be an entrepreneur?

Avak Kahvejian: Grad school.

Liz Pavone: And one goal for 2024?

Avak Kahvejian: Really establishing my PBU (Pioneering Business Unit) and developing my Origination team.

Liz Pavone: Thank you so much for being here, Avak.

Avak Kahvejian: Thank you.

Liz Pavone: For more on Avak, the companies he's helped to found, and Flagship at JPM, visit flagshippioneering.com.