a16z Podcast - a16z Podcast: Connecting Hearts, Bodies, and Networks to Cure Cancer

Episode Date: January 26, 2019

Veterinary oncology can inform human oncology, and vice versa -- providing a better model for looking at drug performance, interrelationships, and more. Especially when you add in data (there's no "doggy HIPAA!") and networks to get a "living laboratory at scale". Or so argues Amy Abernethy (Chief Medical and Chief Scientific Officer at Flatiron Health and advisor to One Health), who was recently named the new Principal Deputy Commissioner of the FDA, pending ethics clearance; and Christina Lopes, CEO and co-founder of One Health; in conversation with a16z bio general partner Jorge Conde. Dogs -- as a species, as pets, as companions, as family members -- evolved alongside humans, so are actually more similar to us... not just genetically and in terms of the biologic pathways that may cause cancer, but also in exposure to similar environmental factors as well. But what does this all mean when it comes to thinking about real-world evidence in science, human clinical trials, and more broadly, building a bio company? How can product designers -- of all kinds -- backwards-architect their product roadmap for data network effects? And how can bio founders keep both a big-picture roadmap in mind while also focusing on specific milestones, and while working across unconnected disciplines as well? We cover all this and more in this special episode of the a16z Podcast, recorded during the recent J.P.M. healthcare conference in San Francisco. The views expressed here are those of the individual AH Capital Management, L.L.C. (“a16z”) personnel quoted and are not the views of a16z or its affiliates. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by a16z. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any securities or digital assets are for illustrative purposes only, and do not constitute an investment recommendation or offer to provide investment advisory services. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund and should be read in their entirety.) Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by Andreessen Horowitz (excluding investments and certain publicly traded cryptocurrencies/ digital assets for which the issuer has not provided permission for a16z to disclose publicly) is available at https://a16z.com/investments/. Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Past performance is not indicative of future results. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see https://a16z.com/disclosures for additional important information.

The content here is for informational purposes only, should not be taken as legal business tax or investment advice or be used to evaluate any investment or security and is not directed at any investors or potential investors in any A16Z fund. For more details, please see A16Z.com slash disclosures. Hi, everyone. Welcome to the A6 and Z podcast. I'm Sonal. So today we have another one of our special episodes with more special guests that was recorded during the recent JPM healthcare conference in San Francisco. The topic is all about diagnosing and treating cancer in pets to humans, but we also go into other topics relevant to all kinds of health care and tech startup founders,

from data network effects to real world evidence and science, to clinical trials, to how to think about product and roadmap when building a biocompany. So first our special guest is Amy Abernethy, who has been named the new principal deputy commissioner of the FDA, the U.S. Food and Drug Administration, pending ethics clearance. She is currently the chief medical and chief scientific officer at Flatiron Health. And we also have Christina Lopez, CEO and co-founder of OneHealth, which provides dogs access to the latest cancer treatments and in doing so contributes data to developing better future therapies for humans. She's the first voice you'll hear. And finally, we have A6 and Z general partner Jorge Condé of

the BioFund, who will be moderating this episode, which ends with the question of what does it take to build biocompanies, but begins with a question about why cancer and dogs and the link to humans. You know, dogs come from wolves. They co-evolved with us, right? And similar to other animals, they are predisposed to cancer, as we are, very similar. The idea there is certain litters that got bread, sadly got bread with their beauty and, you know, all sorts of great things. And sadly, cancer also, which is a disease of genetic origin and mutation. Moreover, the epigenetic factors, the dogs, as they shifted from, say, farm animals to your home, they're also sharing, often they eat the same food you do or eat the scraps

of what you eat. Obviously, they don't smoke, but they're exposed to secondhand smoking. They don't drink alcohol, so that one we can take off the table. But there's a lot of the environmental factors, and then if you factor in, things like they're lower to the ground, they're near golf courses, right? There's a plethora, and that's a more complicated, you know, subject. I would have never thought of that. Yeah, yeah. There's actually studies about that. Certain breeds near golf courses have had actually bladder cancer, transitional cell carcinoma. It's actually an area of study. So it's actually

a very compact way to look at, you know, a snapshot of disease and health, hopefully also, with those who give us their best, right? Has the dog genome been sequenced? So the genome of the dog was sequenced, but very recently, I believe it was, you know, 2010, 2005. It's very recent. Veterinary medicine has not had the resources. And even from the NIH level, what the mouse has been very well characterized and studied, the dog really hasn't. So for me, as a founder, I see a massive disconnect between the importance of the dog, the growing importance of the dog in our lives.

lives as a source of love, and then the lack of funding. Yet, the families, as you rightly put, are spending and it's cash and it's out of pocket and it's expensive. So let's talk about that. Six million cases a year of diagnosed cancer in the United States, in dogs. How do they get diagnosed and how do they get treated today? So the diagnosis is a histopath. So that's basically it stops there. They said that tumor tissue. To do histopatology, that means you, you've gone to your vet and you suspect that your dog has cancer. Lumps. The dogs will start lethargy, doesn't want to eat is a big thing.

This I hear a lot is my dog has come to the door every day and over the last week, not once. That's a, that's, you know, radical change of the behavior. And from the data side, what we're very interested in also is capturing the data from the families who, this, their studies also, who will neglect themselves, you know, If you ask them to take diabetes medicine, they don't take it. They don't give you any feedback. But we'll be copiously observing the dog and will give you huge, like, annotated reports of like, you know, fluffy, didn't eat. You know, she loves this type of food. She usually, you know, runs to me at the door.

You know, all the little activities are not happening. So that's a big, that change of behavior is a big one. And actually what we can capture there as disease progresses. in real world evidence is really fabulous. So, Amy, you've been a long-time advisor to One Health. What is the link between dog cancer and human cancer? Well, you know, it's really interesting. So first of all, I'm a melanoma doctor by background.

So I was a human oncologist, definitely not a veterinary oncologist. Do dogs get melanoma? Dogs get melanoma. They certainly do. And to be honest, before about five years ago, I didn't know that at all. My interest in dogs and sort of veterinary oncology came, as I think, started to understand the link between veterinary oncology and human oncology and how understanding better and being better able to take care of pets could also help us understand

and be better to take care of humans and vice versa. So what is that link? What is the link between dog and human oncology? So when, you know, Christine was talking about all of the different aspects that are contributors to a dog developing cancer, some of those are genetic and genomic aspects. The fact that there's a lot of inbreeding, some of them are exposure issues, so the fact that there's secondary smoke inhalation of the dog or exposure to pesticides. So as we think about the biologic pathways that ultimately lead to the development of a cancer, which is essentially cell division gone awry, those pathways are very similar between the dog and the human. And further, not only are some of those pathways similar,

are actually at the biologic level within the context of the cell, their pathways are very similar. Is that more similar, say, dog and human versus mouse inhuman or primate in human? Is there a reason why dog is a good model? So there's much more conserved biology between dog and human than mouse and human. And as a matter of fact, this now becomes a better model to think about, for example, exposure to a anti-cancer drug, and how that tumor is going to now melt in response to that drug or not. And so ultimately, if we think about human oncology informing veterinary oncology or vice versa, it is a better model system to look at the interrelationship between drugs and whether or not the tumor is going to go away.

So this is, I mean, this is, for me, this is not common knowledge, that there's such a tight link. What was the aha moment for you that made you connect those dots? I sit on a component of the National Academy Medicine, sometimes called the Institute of Medicine. There's something called the National Cancer Policy Forum. And a couple of years ago, we had a meeting on the conserved biology between animals and humans and specifically dogs and humans. And for me personally, the aha moment was I was sitting in this conference, which frankly I was against having in the beginning because it seemed like it was. was extraneous. And I'm sitting there. And the fact that if we could better understand how drugs perform and their impact on the tumor in dogs, we could accelerate our understanding in

humans was a huge aha moment for me. So do we test human drugs in dogs as part of the approval process? Not standard. So it's not necessarily standard of care. The whole message that really I walked away from that conference with was that if we now embedded into the drug development process for human drug development, a better understanding of how those drugs performed in dogs, we could save many tens of millions of dollars in the development of each drug just by having better go-no-go decisions because we know how those drugs are working. And importantly, this is not just about testing in dogs where dogs are essentially, you know, test model, but we don't care.

But this is because we actually can also understand, do these drugs help take care of pets? And can we ultimately treat veterinary cancers in the same way we can better treat human cancers? Yeah, so full circle. So something that's really come up for us is both from the genomics side, but also immunologically. Dogs that have natural disease, spontaneous disease,

have an immune system also that's working. And that's a radical difference from the mice models that are much more engineered and sometimes don't have an immune system to have the cancer and then the ones that they really try to figure out to have immune system and cancer lasts 20 days.

So it's an incredible challenge in immunolology. So the mouse model is an engineered model to actually force the essentially the appearance of the disease. And dogs, dogs. It's naturally occurring. Natural, spontaneous cancers

that are analogous to human versions. Exactly. Well, of course it's a different species than human, it is indeed much closer to the human than the mouse is. And, you know, in the case of primates, first of all, I think the regulatory framework is, you know, not super becoming on that department. But also, they're used more from a toxicology angle. So you really don't have in certain circumstances, especially as the science gets very sophisticated, a real kind of model. But there is a huge proportion of dogs who get naturally occurring cancer.

That natural activity or that natural biology sets up a living laboratory at scale. Now, how do these two dogs get connected? I had been at that conference. And now I'm sitting in a meeting in Boston. And I literally am talking to the guy sitting next to me in lunch, which happens to be Ben, Christina's co-founder. And he's telling me about this idea that they're going to start to devise you. develop a company to work on dog cancer.

And one of the things that struck me, actually there's many things, but first we were talking about the role of technology and that there would be the potential to develop a linked data system across dogs in the country. We talked about the fact that there could be dog clinical trials and that dog clinical trials could have parallels to human clinical trials. And then we talked about the fact that pet owners actually wanted this and there was endorsement from the national organizations around pets. And then we almost a month later met at a roundtable about clinical trials.

And one of our assignments for the intimate discussion was, you know, like, instead of just complaining about the development, we was very oncology focused of drug development, what is like a real tangible solution that we could do right now. And that was the assignment. That was how we were going to get dessert. And I raised my hand. And I was like, I have one. And I have one, and it's probably in your house.

And I started discussing the role of the pet dog and the impact of cancer. And luckily, I had Len Lichtenfeld, who's from the American Cancer Society. And he was like, oh, my God, yes, I'm super devoted. And do you know the story of rebutinib, a lymphoma drug, that, you know, the dog and the human. And actually the evidence from the dog helped not just this company be alive, but the drug be alive. and he said how his dog had lymphoma and a friend had lymphoma and he was able to somehow interconnect. And I mean, he was almost in tears, you know, and I was like, oh, I'm in the right place. So, Amy, you obviously are at Flatiron. Data is very valuable. I would have never thought of sort of a national or global network of dogs being treated and the data they generated as being an important source of data.

So let me ask you, I guess, two questions related to that. First of all, from a mechanic standpoint, how does that data get? housed. Do dog EMRs exist? Is there such a thing as dog HIPAA? How do you actually generate and aggregate all of that data? So when I was sitting there talking to Ben that day, two of the other aha moments were that he was telling me about dog electronic health records. So they do exist? They totally exist. And there's also centers for excellence and the, you know, key opinion leaders in veterinary oncology, they all know each other. So very similar to the human oncology system and the system of managing human health, electronic health records and humans and how we move science forward with key opinion leaders, that all is replicated within the

veterinary system, except for I don't think we put it to use as well. So that was the first thing that was obvious to me. The second thing was that, oh my goodness, there's no doggy HIPAA. Now, we have to do the right thing by pets and pet owners, and we need to basically only think about using that information in a way that's good for society, for pets and pets owners. But that being said, many of the impediments, as we think about data aggregation and data use, that we run into in human health, we don't run into in dog health. Because the spirit of HIPAA is to protect patient privacy, among other things. Personal health information for people.

And if any dog owner knows, dogs don't worry too much about their privacy. Right, exactly. People are so effusive about their dog stories. And their pets are, you know, hugely a part of the family, but also something that you talk about freely in many different ways, we're much more closed about our personal health information than we are about our pet health information. And so, you know, if I now fast forward to that vision of how this might all come together, it was the realization that at scale dogs were developing cancer and that at scale their information was being collected in common frameworks. that if we put that together at scale, then we could overlay the process of drug development

and the process of learning how to take care of pets better. I work at Flatiron, where what we've got is a combination of electronic health record, as well as the ability to curate and clean up data in order to now put that into action to learn what works for whom and when in the human oncology space. So I can see how that would actually come together. I also said on the board directors of Athena Health, where I could see that, you know, not only do you apply this approach within the context of one disease, but you actually now can aggregate data across electronic health record across health and start to get a almost ringside seat across health care. And so those two pieces in that day at the meeting in Boston were in the

back of my mind, as I was thinking about how this might create a national or international system to try and accelerate drug development both for dogs and humans. So just touching on what Amy is talking about in terms of, you know, no doggy HIPAA, I want to just layer on that there's, you know, the accelerated time frame of disease, of life of the dog, right? So you hear commonly spoken about the seven years to one. And if you correlate cancer with aging, the dogs after the age of even five, the probability of cancer goes way up.

So we can see a lot in a very tight time frame, which from a data angle is very interesting, we really want to transform the story from the dog getting a cancer diagnosis of a death sentence, which used to be for pediatric cancer. How can we get into the lymphoma journey of dogs much earlier and intervene? This is a really interesting question. So early diagnosis of lymphoma and dogs is possible. So two things. One is that having dog owners and veterinarians think about the role of one health earlier in the lymphoma journey as well as any cancer journey is not only potentially good for the management of that dog's cancer, but also will ultimately contribute more meaningfully to the overall system of how we're going to figure out how it better take care of dogs in the future. That's the first part. The second part is similar to what we know from human cancers, the better we can get at early diagnosis. it's this early understanding of exactly the details and features of this particular disease, both at the genetic and histologic level, but also in terms of factors that we're driving

it and how this person is going to perform, there's similar parallels in dog cancer. And as the science evolves, and I mean from the human side, bringing what's top in human, back to the dog at an affordable price. When a dog is diagnosed with cancer, are they treated? Is there chemotherapy for dogs? Yes, there are. The toolbox is probably 13 therapies in total that we were talking about with Cheryl London, one of our KOLs that's really top in K9 oncology, total. And the chemos are quite old for a lot of them. Some are not so bad. But then there's an issue and I want to talk about that because it's really dear to my heart of access. Not everyone can get to a vet oncologist, which there's only about 300 in the United States. And think about six million dogs, three, four hundred, right? In the whole country, there's only three or four hundred vet oncologists.

Yeah, okay. It's a special team, right? How many veterinarians are there? About 23,000, I want to say. So it's a very big difference. So there's a really limited scope of access to top treatment. So we want to help democratize that. And as we democratize, we come in earlier.

And I'm very passionate about access that I've been, you know, on the advisory board of international plant parenthood for many years, so the whole Western Hemisphere. And all I cared about was like providing access to, you know, pre-cancerous vaccines, HPV cervical cancer in Bolivia. I always used to say I want to care for those that are voiceless and on the margin and kind of be a fighter for that with all the technology. So on that note, we're talking about catching disease earlier so we can have, you know, more successful intervention. Humans have cancer predisposition genes, Braca, probably being the most famous. does the equivalent thing exist in dogs as far as we know? Is there a BRCA for dogs?

So it's a much limited, right? Like we are working and developing this data set. One that is well, let's say, well known or characterized is a BRAF in TCC in transitional cell carcinoma and bladder cancer. However, we don't know if it's a driver mutation or not, but at least we know the correlation and it's 85% of dogs with transitional cell carcinoma. or B-Raf positive, and it's V-600E, B-Raf, very particular. But there are a lot of discussions and a lot to learn.

So help me close the loop here. You know, where this conversation starts is a connection between dog health and human health. Christina, you're focused on one health here to really change the way we think about managing and treating diseases in dogs. One of the things that struck me that you said, Amy, was that the dog could benefit a lot from the human. but it sounds like the human can also benefit a lot from the dog. So help me close that loop. The two sides of the leash.

The two sides of the leash. We've been trying to be very intentional about being able to collect information at scale that has detailed information about the dog, about the dog tumor, including histology, and some of the genomic information, and hopefully soon more immunologic information, treatments and outcome. And by being very intentional about the product development,

so that as, for example, the phytocure product, which is a combination of a diagnostic and treatment, starts to now become more widely available, the more use of that product, the more that we get wide-scale information that's got those details about disease and outcome. The reason that's important is that the more that we can understand that interplay in dog, genetics, and histology of disease, treatment, and outcome, we've got the same set of expectations and opportunities to understand that on the human. human side so that we can do a number of things. First of all, we can have very accelerated understanding of treatment and outcome that we can now start to think about on the human side, whether that's in the development of new human drugs. We can accelerate the treatment of dog drugs using essentially rural world evidence type mechanisms. And we can start to think about how do we understand more genomic information and the interrelationship with treatment and start to push that conversation forward. So we have both diagnostic and therapies. They're all

precision therapies. We're taking everything from the best that we could get our hands on from the human. We have the first next gen sequencing panel for dogs. So we've been working on the genomics and this would be the first. Do you sequence the dog's tumor? Yes, absolutely. We've done many that we've done the whole genome. We've done gene expression. I mean, we've done every enchilada that you can think of. Our panel, it's the first of its kind. So I'll have foundation medicine call it. So foundation medicine for dogs. Yep.

And then we make a report. Okay. All right. We're doing our real world evidence research. That's our, you know, scientific R&D hat, really. But we really want to treat and serve the community. So to that end, we look at actionable mutations. And then what therapies are out there that we can then bring to the dog to, you know,

target what we believe might be a driver mutation of that dog's cancer. So this is precision medicine. Four dogs. Taylor made. Taylor made based on the genetics of their cancer. Exactly. And we're also looking by like tumor type also. You know, where can we draw some correlations?

It's not as well known as in the human. But what data there is so that we can also work with perhaps even faster, quicker tests and interventions that accelerate the intervention. So those two. You mentioned that there's the potential as we learn more about drivers of disease, if we learn more about what's effective. Are there any limitations on how you can think about combination therapies for dogs? We're seeing an explosion across oncology of the development of combinations,

including combinations in the immune oncology space. And so think about setting up this network. Now imagine that the dog has two potential targetable mutations. And perhaps we've got early qualitative or quantitative information from veterinary oncology. key opinion leaders that suggest these two drugs can come together. And now we can recommend, within the context of the report, you may want to consider these two drugs together, but here's what you want to think about in terms of dosing and watching for talks.

And then what that allows us to do is give the information back to the veterinarian, who may not be a veterinarian oncologist, who can then recommend this particular treatment for this dog and this pet parent. So you provide the therapy? provide the access. And then we can follow what happens over time. And that will then create a reinforcing data set to go along with what we know from the key opinion leader to now understand better the combination, not just the individual drugs. So, you know, we spend a lot of time thinking about data network effects. And so what you're describing here is sort of a canonical data

network effect that the more dogs you treat in more combinations results in better information to treat more dogs with better combinations and so on and so forth so the product gets continuously better. Is there an ancillary data network effect that actually accretes back to the human side? Is there anything we can learn about combination therapy that might be translatable back to how we think about treating human cancer? I think about it in a couple of different ways. The first is that as all of those sites are networked together, essentially the data sets are networked together, the first thing we can do is start to understand how combinations potentially are synergistic, or those combinations may actually be harmful in combination, and we'll

understand that better within the context of dogs. The second is that network doesn't just have to now be referential to drugs that are already currently available and are sort of passively given within the network. We can start to get intentional and start to look more like clinical trials because we've set the network up to now be ready for more clinical trials-like model. So it's like a virtual clinical trial. It's a virtual clinical trial. who appropriately consent in place. And that sets us up to be able to study the combinations that we'd like to be ready to do in humans,

but we're not ready to do it quite yet because we don't have enough data. Historically, we'd only do it in mouse models, but now we can pressure test that in a dog-based model with appropriate permissions in place from the pet parents and then be able to translate that very quickly to the human space. Is this a flat iron for dogs? This is flat iron for dogs.

Yeah, yeah. We've already begun. to do that, both in the traditional more R&D setting with clinical trial sites, but then also now, you know, thinking from a flat iron and real world evidence and the longitudinal data with the outcomes. But can I make a comment about product development as it relates to this? You know, when I first started talking that day, my mind was already on the clinical trial side. And if you go back and you sort of think about this sequentially in terms of a startup and sequential product development and being able to think through what are the things that we have to

have in place first, right? So being able to have a product that veterinarians today can understand and see as valuable. And so can pet parents, that's a pretty critical step one. Being able to think about what are the unique data points that we're going to need to have routinely embedded in the system to be able to now make that clinical trials model move faster is step two. Being able to be, you know, confident about what drugs to give and how to give advice to veterinarians, a the country is a critical step in this. I think that's probably step three or maybe step one point B. And then another thing that we have to get right somewhere in this process as endpoints.

So we have to be able to be confident that this is how we know that the cancer is getting better, whether or not the dog's passed away because of the cancer, toxicity, et cetera. So as we've been trying to set up the clinical trials model, we've actually gone back to the building of the network and building of the data model to ready all of those things because of that stuff's in play. then the clinical trials model just flows really easily. That's extraordinary, right? Because there are really two products here that are, of course, interlinked. But product one is the benefit to the dog, the treatment of the dog.

And product two is the data that gets generated over time with the experience that all of the animals that go through the network. And that's very similar to what we saw flatter. So what we saw flutter is you've got to make sure that you take care of oncologists and patients first and have a mechanism to bring the data in. And then you can get the data ready to have secondary purposes. But if you have that out of order, you're actually, you're never going to get things. Yeah, well, you know, in marketplaces, there's the saying that the folks use,

which is come for the tools and stay for the network. Yes. Right. It's an element of that. So let's talk a little bit about company building. So, you know, you're going through what is going to be an incredibly ambitious, you know, product development roadmap, a very important one, given the, sort of the problem as you've laid it out and the challenges that you've laid them out.

How do you build a company in that context? Number one, you have to be, you know, kind of courageous and want to take risk because it's not, you're building the train and the tracks almost at once. For us, we've surrounded ourselves with folks who can help us, right? Because the idea is you're going to make mistakes, but you want to lessen the mistakes you make. And that zero to one is a very, very particular world. I don't know if there's a resume for that. So what is leadership is really jumping in that fire and that challenge, but being prudential.

You have to a bit work on your vision, not just like a stare, you know, where there's not established set of rules. So you're going along and you have to be very in it, paying attention to the details and almost at the same time with like a big canvas view. So what advice would you give the founders in terms of, you know, sort of pitfalls to avoid or mistakes to make? So, you know, some of it is what Christina brought out in terms of you need to be able to telescope out, but also then get back into the weeds. I think there are a couple of critical pitfalls that I've seen over and over again.

The first is create some short-term goals and be really, really intentional what you're going to do at three months and maybe three to six months. Six months plus is where I'd like to get to, but don't spend too much of your time and hear the details of exactly what I'm going to do in six months. and so plus until you've got early data coming back. When I say data, I mean feedback from the market and, you know, what your investors are telling you and, you know, what your mother tells you about what she thinks about what it is you're doing, right? Because that feedback is very critical to chart your course. Meanwhile, when you're doing something like when health is doing, you need the input of key opinion leaders and especially scientific experts. I'm a clinical trialist by background. I've been reading outcomes research studies. I've been doing large-scale science

for 25 years. And we have to think five and eight years out. So in that hat, I'm going to come at the problem and provide you my advice with an eight to 10 years out hat. And so as the founder trying to work within this space, you have to think about how do I translate that into my three and six months product roadmap vision and conversation without dismissing it as an important, right? Because that is really critical information.

but they're not, those advisors who are bringing that scientific advice to you, either in your company or as external advisors, are not going to understand why you're not doing all the things that you have to do for that seven-year time frame because really what you've got to do is focus right now. And so I think that the biggest, you know, pitfall that I have seen over and over again is the inability to walk that line and you either dismiss and somewhat disrespect one group or the other somewhere along the way. So that's super helpful perspective to have, given that you guys are on the ground as founders. I think one of the things that I've always found in building a bio company, and this is a unique one because you're outside of human health, you're in dog health, there's going to be a consumer aspect to this. But one of the biggest challenges is figuring out how to narrow product scope given the massive mandate that you have to try to cure disease. And it's inspiring to hear you go through the product roadmap in terms of what you're looking to do. I think you're doing it absolutely in the right order to say product first to the benefit of the dog and the pet parent and then, you know, generate the data and the data, you know, obviously will benefit the network over time. But to me, I think it's one of the biggest questions is, you know, you could always do more with the product.

You could always do better, but remembering that right now what exists is so basic that not overengineering the product, I think it's one of the biggest challenges that exist in the space and finding that right balance between making something that's beneficial. In this case, to the pet and the pet parent, recognizing that it can always be better and it always will get better. And there'll always be more data you can incorporate. There's always be more sophistication that you can have, but not trying to sort of nail that in version 1.0 because category creation is a very difficult thing. So I'm very much on this boat. Internally, folks, they're like, wait, it's not perfect. Wait, it needs to be perfect. What do you mean? You know, it's going to mess up the whole data set if like that is not perfect. Or external or advisors are like, well, I have to see biologic activity and you have to prove it to me now.

You know, so it's really, it's a tough balance and to harmonize having the patients, but also validating what's important over the roadmap, but then what you have to do now, right? There is a first step to that sequence. This is music to my ears, but hard in a bio company because you're also hiring folks that are coming out of school, right, and they're very ambitious, but very, you know, maybe, you know, we have a multidisciplinary team, so they don't, all come from understanding even, you know, the challenge of cancer itself. Well, actually, on that one piece of advice I would give you when you hire multidisciplinary

people, some coming from tech, some coming from the veterinary world, others coming from the entrepreneurial world, is, you know, one of the big challenges you always have is that each group doesn't know what they don't know about the other group. So, for example, Amy Abernathy is an oncologist. There's an incredible amount that she knows about cancer. Someone that comes out of the tech world probably doesn't even know what to ask her. And so one of the things that makes cross-functional teams in the biospace especially important is finding a way to develop shared languages and shared frameworks for thinking about disease, especially in biology, which is messy, because otherwise you run the risk that people are actually talking past

each other without realizing that they're talking past each other. So Amy, you've been named to a senior role at the U.S. Food and Drug Administration. Can you tell me, me a little bit about the role as you see it and what you hope to accomplish? And I'd also be very interested in hearing a little bit about the journey that led you there to the FDA. So this is an incredible honor. As I step back and I think about the things that have motivated me for the last 20 years, I've been really focused on how do we get treatments to patients who need them faster and how do we have confidence in how those treatments work? And also, how do we make sure that the treatments are specifically effective for that person and as specific as possible? I worked on this when I was in

academia. I then kind of hit the boundaries of what I could do from the academic side. So I thought, okay, I'm going to try and go and do this from the private and commercial side. And so I went to Flatiron and I was on the board of Athena Health and did other things, including One Health, and have not hit the boundaries of what I can do from the commercial side, but I think what it also did was give me a visibility into what potentially could be done from the government and regulatory side. And so for me, going to the FDA was this incredible honor to be asked to do it, that also lined up with this natural progression of trying to move that forward. When I'm at FDA, I will have a trans-FDA remit, so I will need to spend my time.

and energy working on the things that the FDA needs me to do, and I'm looking forward to learning more and more about that every day. But I know that I'll at least spend some of my personal extra energy on three core areas, one being accelerating evidence generation. I can't work with One Health anymore, but you can sort of hopefully see how I've been thinking about evidence generation for a long time. The second is as it relates to personalized medicine. This has been a core focus for me for a long time as an oncologist and that's just as a person. And then the third is patient-informed drug design. Extraordinary.

And as a citizen, the United States, I'm grateful that you will be in that role. Thank you very much. Me too. Thank you, Christina. Thank you, Amy, for joining me today on the A16Z podcast. Thank you both. Thank you. Thank you so much.