The a16z Show - Building the First CAR T Company

Episode Date: February 7, 2020

with @OzAzamTmunity1, @JorgeCondeBio, and @omnivorousreadCAR T therapy, the groundbreaking new medicines that uses engineered T-cells to attack cancer, has been so effective in childhood leukemias that we believe it may actually be a potential cure. But this isn't just one new medicine, it's an entirely new therapeutic tool—and a total paradigm shift from most traditional medicines we've seen before.Tmunity CEO Usman "Oz" Azam was previously the head of Cell and Gene Therapies at Novartis, in many ways the first CAR T company and the team brought us blood cancer CAR T-cell therapy Kymriah—the first cell-based gene therapy to be approved in the US. In this conversation, Azam discusses with a16z's general partner Jorge Conde and Hanne Tidnam what CAR T therapy really is and how it all works. The conversation begins with the “patient and cell journey” of this treatment and how this medicine is developed, manufactured, delivered to patients; why exactly it's so different traditional medicines; what it will take to make these new medicines work on more kinds of cancer, scale to more patients, and cost less; and finally, what company building lessons can be learned from building the first CAR T company of its kind from the ground up.This episode was recorded at the annual a16z Summit. Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that 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. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Hi and welcome to the A16Z podcast. I'm Hannah. This episode is all about the new medical paradigm of CAR-T therapy, a new cancer treatment that uses engineered T cells to attack cancer and has been so effective in treating childhood leukemia's, we believe it may actually be a cure. In this conversation, Timmunity CEO Oz Azam discusses with general partner Jorge Condi and myself all about what car tea therapy is and how it all works, starting with the patient and cell journey to how this medicine is developed, manufactured, delivered to patients, how different it is to traditional medicines, and then what it will take to make these new treatments work on more kinds of cancer, scale to more patients, and be more affordable. And finally, what company building lessons can be

learned from having built the first car tea company of this kind from the ground up? This episode was recorded at the annual A16Z summit. We're here to talk about this new kind of therapy, KART-T therapy, and what it means to be building a company that is delivering this brand-new medical paradigm for cancer treatment. So let's just start by giving a little bit of background. What is your tagline of here's what KART-T is? So KART-T is also known as Kimeric-antigen receptor therapies. Nature's biggest gift that we were given in terms of protecting us from disease is something

called T-cells. There are a subset of your blood cells that are called white cells. White cells typically prevent infection disease, so they are always surveilling and protecting you. A B-cell produces antibodies. A T-cell actually hones in and gobbles up peptides and abnormalities that are circulating in the system. And the idea was, could you combine the features of a B-cell and a T-cell together? And that's where the chimera comes in. So, chimera was an ancient group. Greek mythological figure, right? There was a hybrid, I think, of a female lion, a dragon, and a serpent, or something about nature.

Yeah. So the whole idea being, could you combine and create a blend of something with the idea that you could create therapies around it? And the nub of the therapy really involves taking a patient C cells and we re-engineer those T cells. Think of it like a GPS system in cells that we've been able to engineer. We take cells from a patient. We engineer them, we give them back, and those cells detect cancer and destroy them.

Best analogy I can give is like a SIM card into the T cells. That SIM card that gets expressed on the surface of those T cells is very unique. It only dials one number. And that number is a specific cancer antigen or a protein that's an abnormal protein on the surface of cancer cells. And we're able to get these T cells to then actually become killing machines in some ways, whereby they identify an abnormal protein on the surface of a cell, and they go and attack. So let's do what I call the patient journey and the cell journey.

So I'm going to take a profile of a child with leukemia. You have a child of the age of three or four. They start getting bruising. They go to their family practitioner. They do a CBC. They look at their blood count, and they have massive leukemia in terms of their white cell elevation. The child gets rapidly assessed. They start chemotherapy in great news.

they respond. And most kids with leukemia respond really well to chemotherapy. Two years later, there are a routine follow-up and boom, the next thunderbolt comes in. Unfortunately, they're starting to now get leukemic breakthrough. There's more chemotherapy's provided. But then there comes a point where these patients become, what we say in the oncology world, refractory relapsing. So they're refractory to any further chemotherapy occasion being given to them. And they're relapsing because their disease is worsening. And so that patient is then brought in to have their blood drawn to see, do they have that right surface marker that you could create this engineered therapy for?

If they express something called CD-19, then we basically harvest out their T-cells in a process called aphoresis, whereby patient's blood is withdrawn through a machine and it filters out the white blood cells. Those cells are then taken and they're shipped to a central manufacturing facility in the case of the University of Pennsylvania, they actually have their own manufacturing capability. So they do it on site. They do it all on site. And remember, this patient is sick. Yeah.

So you've harvested their cells. Yeah. You then go through a process of seven to ten days where you have to re-engineer those cells. Those cells go through a process of cell selection, sort of right cells are extracted. They're then excited by a certain degree with certain technologies that basically make the cells in a receptive state that you can then deliver a Trojan horse into it. The Trojan horse is this payload that we deliver of the genetic code that expresses this new surface marker called a cart on the surface of the cells. You then go through a process of three days watching these cells, are they going to grow? And you know, you cross your

fingers and toes because sometimes they don't grow. You know, these are cells that have become fatigued and they just don't have that oomph, that energy that's needed to grow. Then you have to harvest out the cells once they've grown. Then you have to freeze them. Then you have to ship A mild chemotherapeutic regimen is given to the patient. We kind of call it conditioning. And conditioning is that you want to get the patients in a certain state that you create space in their body for them to receive these cells and the cells to expand. Interesting. So the cells are given as one infusion.

And what you typically see is a spike in the patient's fever. These cells start to multiply very, very rapidly. And at the same time, they're pushing out massive amounts of protein. And they start to literally attack the cancer wherever they see it. Cancer, when it's destroyed, releases a lot of toxins. And that manifests itself in something called cytokine release syndrome. It's like a storm in your body, right? That's what they call the cytokine storm.

And so having that patient available to be able to, for example, move to an ICU unit if needed, it requires a lot of coordination and sophistication, right? So you then go through that process and hopefully by three, four days, you're seeing that window of, is this patient really responding? If you don't see the cytokine storm, it means the product's not working. Interesting. We actually look forward to an adverse event, which is really weird in medicine. Yeah.

Because if you don't see it, you know the product's not working. 28 days later, when the patient is better, the fever subside, and you do a bone marrow biopsy, you do various blood tests, and you see over 90% of kids initially in the trial's got complete remission after 28 days. And there are children out now, you know, seven, eight, nine years now. And that is the cure. That's persistent and durable cure. We hope that they remain in this state where these cells are constantly in surveillance in the body. So should a signal a rise of an abnormal protein, these cells can then attack it.

So I've given you a sense of the cell journey and the patient journey. Now you think about that, creating a product around that. It's a whole new area of medicine, right? The infrastructure, how do you begin to scale a process like that, to build the pipes and the infrastructure to scale? If I go back to 2013, literally, we'd be in the size of a room like this podcast room. Right. And literally we would have tubes and bags hung on the wall. It was literally our sort of brainstorming war room of how do we take this process from an academic, open process, close the manufacturing, meaning lock it to good manufacturing practice of standards.

Process development, analytical development, vector scientists, and technical operations personnel are working around the clock. So, again, a very different way of practicing medicine, right? Yeah. This was like the Wild West in some ways in the early days. But we did it. And we learned a lot through that process. We acquired our own manufacturing facility because we're not in the business of just creating product for chronic sake.

We want to actually expand it globally. We need to bring down the cost of goods radically for these therapies because they are really expensive to make. So unless you invest it upstream in there, then how are you going to be able to scale and actually make these products affordable? and at the same amount of time, you know, generate revenue for the company. Yeah.

The process is so important. It's so different to traditional medicine. So you have to be able to manufacture this therapy. You've got to be able to manage the logistics that go from patient to the provider, from the provider to the manufacturer, back to the provider, back to the patient, what you call the vein-to-vane logistics. So is there really any other way to do this but to be a full stack? or fully vertically integrated company

if you're going to commercialize these types of therapies. I think the more and more you see where the world is moving to and you look at the personalized nature of what we're doing, whether these are current generation products or off-the-shelf products in the future. That ecosystem being understood from the patient journey, the cell journey, cell logistics to your point, adverse event management,

and as you think about the interface of tech for the future, which is going to be required here, whether that being diagnostics, whether that being management of patient, patient selection, or whether you're looking at blockchain, for example, in terms of secure chain of identity. Because look, if I'm taking yourselves, you want to guarantee I'm giving you yourselves back, right? So there's a whole security apparatus in this and that people just don't consider when they first get into it. If we didn't have that pillar of manufacturing, if we didn't have the research engine, if we didn't have the ability to learn from each patient that we manufactured,

what's working well. Do we need to add a bit of this reagent? Do we need to stimulate the cells in a certain way? All of that, repeat learning. That can only happen in a full stack company. In order to be able to really maximize and create great products, we decided to own that process ourselves. So can you imagine that if we see success in a clinic and we don't have the manufacture to go in hand, I kind of feel that's unethical. You know, in terms of the breakthrough speed with which science is evolving, but not being able to manufactured the product would be such a shame. Building this new kind of technology, this new kind of medicine, the talent, the culture, and the platform. Everything new, essentially. That sounds really painful. It was not easy. It was actually developing products in a different way against the paradigm.

So in our world, you know, of drug development and product development, you know, there's a very well-established cycle of how you do things. You know, it's memorialized with the FDA, there's guidance, et cetera. But try developing something that regulators have never done before, or companies have never done before. In my career, I never thought I'd work on something that could be curative. You know, I worked on things that could help people, they could improve their health. I worked on many things that didn't do anything for patients. You know, products failed. But, you know, once you've touched success in terms of curing a patient, and I use that word very carefully because as a physician, you always think twice about, you're really curing somebody.

When you've done that with a product, it changes your whole perspective about medicine and where the world could be. If you have the right team behind you, if you have the right culture behind you, and if you have the right platforms and technologies. So you had 400 people within the Cell and Gene Therapy Unit in Novartis. You're now the CEO of CEMUITY, your own startup. So Team Unity is a T-cell engineering company that's focused on curing cancer. And we're doing this by developing therapies in the form of either Cart-T's or T-CRs, T-Cell receptor technology. and mainly going down the road less trod when it comes to the tougher kinds of cancers that are mainly in the solid tumour space as opposed to hematological cancers, which there

have been great successes in, still an unmet need. But there's an even huge, huge bigger unmet need with patients with solid cancers, which is where we really want to focus. So taking what you did for blood cancers essentially and moving that to solid tumors. And pivoting from that and the lessons that will learn and very, very important lessons, taking that into how do we develop the therapies for patients who have no other choices left. Can I ask your question around the technology and how you're going to build out your product pipeline?

Sure. A lot of that's going to hinge on how you can essentially engineer cells for increased and expanded functionality. Sure. So one simplistic way to think about this is, you know, in CAR-T, you know, it's an oversimplification to compare it to software, but I will. You know, every generation of CAR-T is built on the previous generation in some ways. And you can swap in, you know, modular components. Cassettes, modules, whatever, yeah. Exactly, to increase the functionality so you can go across different and more complex cancers.

So going from, you know, liquid tumors like the lymphomas, like the leukemias, into solid tumors. I want to talk a bit about how you think about what needs to happen for you to bring the cost down. What has to happen from an engineering standpoint for that to happen. Yeah, so let's break it down. So how do T cells actually bind or stick to a target? Think of Velcro, right? So when Velcro attaches, the idea was that what you actually want is that Velcro never to come off. It sticks permanently really well.

And that was known as high affinity, especially in terms of antibodies. And people realize actually that's not such a good idea because you actually get off target effects. So you actually get something sticking where it shouldn't. There's a problem. So then this maturing of something called affinity tuning. happened where it's the Goldilocks thing, right? Not too hard, not too stuff, just the right amount where you get a T-cell touching and activating, but not totally binding. I think that's been a bit of revolution in terms of T-cell engineering thinking. Then it's actually developing

multiple warheads. So actually targeting more than one protein, abnormal protein, or antigen, that multivalency, as we call it. And in certain cancers like glioblastoma, we know we probably have to attack three, four different kind of surface marker proteins for patients to get a benefit. Then there is the engineering component of getting these cells to power on. So if you think of the engine of the cell, how do you really give it more choke? So these cells really power up and co-stimulate? How do we armor these cells better to help them overcome immunosuppression?

Because Mother Nature's, you know, one of the things that she's done an amazing job of is actually giving cancer privilege, right? And there are various ways in the system that cancer cells trick the human body. One of them is immunosuppression. So we now have the capability through gene editing, for example, to overcome that or through armoring of cells, which was not possible before. There's actually ways where you can target the outer casing, if I can call it, of cancer, and kind of make a dent in that armor to allow payloads to go in. So payload delivery is another key factor, right, that's changing the way we think about liquid to solids. So all these modules now, the lessons learned from hematological

millingcies now pivoting towards solid tumors. These principles are going to be really important, not just for companies like us, but for the entire field. How do we get these therapies into patients earlier in their cancer journey if and when it's appropriate for them to be treated with these kinds of therapies? And what it needs to happen from a product standpoint to enable that? This whole sort of scale business, that's really going to come where gene editing, comes into play. In theory, if you have not the patient cells, but cells from multiple sources,

so you're a donor, I'm a donor, we have great cells, those universal cells be made into sort of batches from which you can then create an off-the-shelf product. That's going to help reduce the cost of goods, and that's where the future will move to. And there are companies that are actually in that space now. We're in the research stage of that now. But again, that's going to be another flexion point for the field in terms of bringing out costs. Novartis, Kim Raya, the first Cartier product, was going after leukemia's, lymphomas, which is about 10% of all cancers. You're going after, you know, the other 90% plus.

So you're going after call it 10 times the market with a 10th of the team today. Is there something fundamentally different about how you build a startup company from scratch versus a startup company within a large company that you had in Novartis? Oh, there's huge differences. Let's talk about the talent. So I quickly brought on board a very good colleague in mine from Novartis, Michael Cristiano,

and he is the resident deal maker when it comes to sound, everything's salm gene therapies. So between the two of us, we'd actually had a cultural sense of what we had built previously and the essence of what we needed to keep, but what we needed to pivot to. And then clearly the speed with which you have to recruit is a very different pace, right?

I mean, within the Nevada's world, it was fast because we went from two people to within six months having 400 people. But literally we lifted and shifted groups out of different functions in. Sometimes they had a choice. Sometimes they didn't. But in the startup world, you really are relying on your network very heavily, but also really digging in quickly about people who fit really well in big pharma may not necessarily work out in startup situations in small biotech. So really understanding what is the motive of that individual?

I mean, we get out of bed thinking about T cells, we go to bed at night thinking about T cells. Yeah. Well, to your point also, in the startup, everyone is choosing to be there, and they must be making the choice from a kind of passion. You know, many colleagues who came on board had experienced malignancy themselves, or they'd lost a loved one or loved ones in many cases. So it becomes very, very personal. In a small company, you have to phase and stage your hiring.

You can't just do a scattergun approach, right? Getting a high-performing team set up quickly where you buffer each other's weaknesses but play to each other's strengths. Right. That's a really, really important trait in startups because there's no room for insecurity. You have to have huge self-awareness

about what you know and what you don't know. And time's not on your side. It weighed heavily on me the letters and Facebook postings I got from patients and their families in my former life. The patience we needed to treat have tried absolutely everything. The sense of urgency was there as well. For the first 12 months, it was all we needed at that time to really secure the operations of the company. There were many times actually the company should have died. For example, there was IP that maybe wasn't what we thought it was before.

You know, there were platforms that maybe the experiments weren't reading out the way they were. they thought they would. It all worked out. But during that phase, I never knew I could be so tenacious because I tasted such huge success in my former life to now do this again and think, oh my God,

could we really fail? And we really could have. But it's patients awaiting, right? There's nothing like famine to make you hungry. Oh, absolutely, absolutely. And then the next phase of the people side really came in when we had to instill discipline process, building our quality systems,

starting to write protocols, starting to really gear up for regulatory guidances, submitting I&Ds. And sometimes you get people on the bus. Then you get to figure out, are they on the right seat on the bus? Because you hire utility players, right? But you also need specialist players as well. So how do you move them around the seats on the bus? Like, should they have got on the bus?

Hopefully yes. Then sometimes you realize, well, you know, this person may be better in this seat versus that seat. So we went through that phase as well. And did you find that kind of internal cultural building? I mean, because this is a new kind of bio company, really, where there are kind of internal cultural issues as well, where you're getting different types of people

with different approaches to things? Absolutely. People bringing in different baggagees, right, from a cultural perspective. The teaming in our world is so different because you have to integrate what is traditionally clinical thinking with manufacturing science thinking. In our world, the process is the product, right? That's a very different proposition.

So let's fast forward to the end of the journey then. So let's assume that you're able to demonstrate that these therapies are effective. You're able to manufacture them. You're able to manage the logistical complexities. If we look at the first generation of Carty therapies, is it fair to say that commercially they haven't lived up to the expectation of what they might look like? I mean, first of all, introducing a new order in how medicine's practice is not easy. I mean, you've got to think about how a physician's going to get paid.

how does a healthcare system make money out of this? How do you, what's the cost of infrastructure build? Are they going to actually invest in the stem cell lab? Are they going to invest in the logistical wiring? And at the same time, do clinical trials and also be a commercial center. So I think people underestimated the complexity of what it would take. But at the same time, nobody's disputing the stellar clinical results that you get. I think the curves will pick up.

It just took longer for the uptake. for a specialty product to be introduced. And, you know, hindsight, maybe there were certain things we could have done a lot earlier. There's a finite amount of resource investment you could make at certain pivot points within the life cycle of a product. But I'm pretty optimistic that the players

that are going to be in this space are going to double down and increase the actual spend that's needed to really make these products successful because at the end of the day, none of us are in this business to break the healthcare budgets of any country. We also know we have a challenge

because the products are expensive to develop. I think over time, and we've already seen this, the cost of goods have come down. We're one or two engineering steps away from some radical optimization of these products. We are going to need access to large amounts of data. We are going to need the AI-adjacent thinking to be blended into the cell journey

and the patient journey and the commercial journey. The revolution started, it ain't over yet, and there's a couple more wins that I think we're going to see in the next couple of years in the T-Sull engineering space. Jorge, you also built a company where it required you to kind of like both create a new ecosystem and enter the existing ecosystem. What do you think is your primary tool when you were doing that in this particular space? How did you navigate that?

When you think about new modalities, new therapies, you know, biotech-based products, the big challenge is exactly as you say. I mean, just the acceptance of, first of all, clinical acceptance, which should have a high bar in terms of what people think of as a new modality that's worth pursuing. doing. That should be a high bar and it is. But the reimbursement bar is also incredibly high for any new modality because the costs are obviously additive at some point. That tipping point, it does take longer than you think it's going to take. And for sure. But when it arrives, then the dominoes do start to fall quickly. And so the one question to you as it relates to the cell therapies is one of the things that most new therapies don't have is the potential to demonstrate something that starts to look like a cure.

And so, you know, because you had sort of these incredibly outsized effects for the first generation of car T, I would have thought that reimbursement would have followed pretty quickly because you're actually seeing children get cured. It's a really good point. So I think for the leukemia's for children, it was a no-brainer. It happened if you think about it relatively quickly in the U.S. Because it was so compelling, right? So, you know, what's your standard of care? Your stand of care is a stem cell transplant for a kid with leukemia. So cost of a stem cell transplant, believe it or not, is 750,000 to 1.3 million, depending on your zip code.

Okay, when you've fully loaded costs, you add it all in, right? So you are pricing a product at 450,000, 425,000, Kase of Kim, Raya. And after rebates and discounts, whatever that is, right, you can figure out what that number's going to be roughly. You're still getting pretty good value when you think about it. I think the biggest thing for the payers was, and still remains, for cell therapies or gene therapies, you know, anything where there's a price tag that is higher because of the cost of manufacturing, for example, is that don't burden us with a one-time upfront cost for your therapy.

We understand that you've spent a lot of money in R&D. But this whole notion that we're going to be able to discharge our cost up front as a big ticket item, I think that's a thing of the past. And there's precedent now for that. And you've seen presence now with innovation coming through in terms of reimbursement, you know, pay for performance kind of models. You're going to see, I think, variations in that and versions of that. And at the same time, the cost of goods of these products is coming down, be gene therapy or cell therapy. It's a math problem, right?

Engineering problem. And the next five years, that'll improve. But payers, you know, if you look at a medical director in Aetna, United, whatever, they have a finite pot. They want to know in four, five years time, what am I going to be dealing with, right? they're getting so much better now at engaging with companies like us to forecast and think from an actuarial perspective, right? How am I going to manage a business in this area? So part of it is owners and companies like us to be progressive and be creative and have the

engagements earlier, you know, because it is a brave new world. Nobody has the magic answer here. So the earlier you get into dialogue with payers and advisors in that setting, the better off we're going to be for the sake of patience and for the company as well, longer term. So the first generation of Carr came out of Novartis. The second generation of Car T is coming out of companies like Team Unity. There was a reason why the second generation didn't come out of Novartis. In other words, that you started a new company to go after sort of the next horizon.

How do you make sure that the third generation of Carr comes out of Team Unity and not out of a new co that hasn't been imagined yet? curiosity, right? You have to have that, your bedrock as a company. Where is the world moving to? Where is the next best idea coming from? It's how you think about staying ahead of the curve and building that network. And, you know, opportunities arise in amazing scientific settings where you least expect them. I think also people in our world who are known as cell whisperers who really have seen there, been there, done it all, and have seen what works, what doesn't. Institutional memories a really important thing.

They've got founders like Jim Riley, Carl June. You know, they first started off this field in HIV. They were actually trying to find a cure for HIV with, you know, T-Sars. They stumbled upon decoding the HIV virus and realized that this can be a great payload. And look now, we're actually curing cancer with a denatured approach to HIV. So the craziest ideas come from the craziest parts of science in the world, right? Who would have thought, you know, at the time that that would be. be a path that 30 years later would be potentially a cure for cancer.

So last question, takeaway for our other entrepreneurs and founders. You've now twice built a product and a company that's really pushing all kinds of new limits and regulatory and policy and manufacturing and delivery and all kinds of things. What's something that you would do differently now if you were going back and doing it all over again? I think if I had my time again, there are certain things I would have accelerated in terms of my advisors. don't wait for an issue to arise before you think, hey, you know what, I got this problem.

I should try and find somebody that has an external lens to this. And it's where the humility piece has to come in. You cannot know everything. Even your team cannot know everything. So then how do you appoint the right directors to your board who bring different skill sets and advisors? I'd accelerate all of that thinking three times faster than I'd done it the first two times. So agility, tenacity, again, not giving up. the network that you have to create as an entrepreneur in this space to really think good ideas can

come from anywhere and just staying curious constantly. I mean, the one nice thing about our field is we're competitive, but we're also very, very collaborative as well in this space. More so than you will see in other bi-tech spaces because there's a humility as well that we just cannot solve everything ourselves. We will need to beg, borrow, partner with others if we want to be successful for the sake of patience. You can't do it all yourselves.

So humility, curiosity, and the right team. That's good advice. Thank you so much for joining us on the A16Z podcast. Thank you.