Podcast: Adaptive oncology with Dr. Jack Kreindler
Like fingerprints, no two cancers are exactly the same. And yet cancer care – especially for those in the most advanced stages, and with the most complex cases – is administered according to standard, linear protocols. Clinical trials, too, rely on examining one course of treatment against another without modifying the trial based on actual patient outcomes. Cancers are dynamic. They build resistance. They adapt. So why don’t we adapt cancer treatment for each individual, each tumour?
In this Mosaic Ventures podcast, Juliet sits down with Dr. Jack Kreindler to talk about adaptive oncology, the challenges of putting it into practice, and the opportunities for startups to address the massive unmet need in cancer treatment. Dr. Jack is a leading expert in health optimisation, co-founder of the Centre for Health and Human Performance, a serial entrepreneur, and our very own Venture Partner here at Mosaic.
Juliet: Hello everybody, and welcome to the Mosaic Ventures podcast. Today we are joined by Dr. Jack Kreindler who, among many things, is one of our Venture Partners. Thanks for joining us, Dr. Jack.
Dr. Jack: Thank you as ever for having me!
Juliet: Our topic today is adaptive oncology. And as a starter into adaptive therapy it would be helpful to hear a bit more about your work at CHHP.
Dr. Jack: Sure, so CHHP, my institute in London—The Center for Health and Human Performance—is effectively an adaptive therapy clinic and research center. We’ve obviously started looking after elite athletes and human performance science and elite sports medicine but it turns out that treating cancer patients like athletes, using the same principles and methodologies that you do for elite sports people, looking at their exercise physiology, their nutrition, their sleep and recovery, and, to an extent, their cognition and their headspace, it also helps get people with the toughest cancer cases to get through their treatment, and also enjoy better outcomes.
As we do with elite athletes, we’re constantly adapting therapy. We’re testing a lot of things. We’re measuring those factors. We’re coming up with hypotheses. We’re putting together combinations of interventions, be it exercise or so on. And we’re re-measuring, constantly, and re-jigging, and adapting. And it’s something which we do with athletes, when we’re flying airplanes, we’re doing it when we’re making bets on stock markets! We’re constantly making hypotheses, measuring lots of things…monitoring…changing…
But it’s not something we do an awful lot of in medicine. Perhaps if you’re in intensive care, fair enough, where you’re constantly being monitored and you’re wiggling and jiggling things, but that;s not really the way in which evidence based medicine is constructed. It’s constructed in quite controlled A versus B trials where you’re persisting with one thing versus another thing, and then the end of sometimes a very long period—maybe five or ten years—you say, what are the outcomes? And is that new drug better than the old one, or better than placebo? You don’t really adapt in real time, you stick with it.
Clinical medicine, obviously, allows you to sort of see what the side effects are, is it really working, is it not? But largely speaking we’re not doing real time adaptation.
Our thesis is that if we do the same thing as we do for elite athletes when we're training them, or the same things we do for airplanes as we're flying them, we have general guidance as to which direction to go in. But we should constantly monitor and adapt what we're doing in much more real time in cancer therapy. I think that this is giving birth to an area of a new kind of trials called adaptive trials that move people from one arm to another arm to a more optimal arm in much higher frequency than we would we need versus waiting five years to get sometimes historically very bad outcomes for lung cancer trials. But what we're doing is we're moving people towards arms of treatment that look like they're doing better for other folk out there.
Juliet: Before we jump into clinical trials, which is actually going to be a super helpful analogy, let's take a step back. Can you talk a bit about adaptive therapy in practice? There is a lot of buzz about genetics informing disease treatment for example but obviously it's more complicated than that. How does this actually work? And is adaptive therapy actually better than treating people with whatever therapy has been most effective historically? Is there proof of that?
Dr. Jack: What do know is that cancer is a largely a genetic disease. It's a lot of genetic diseases. It's also lots of metabolic problems and a bunch of other stuff and structural things.
But let's just take it as a sort of a multiple genetic disease. It makes sense that instead of just treating bad cells as a whole, I'm saying let's try and stop them from growing.
Let's look at the individual reasons why they're growing in weird ways and not stopping and evading the immune system and so forth. And let's try and target those things.
And that's what I call personalized medicine, stratified medicine, precision medicine. It's not necessarily the same as adaptive medicine. And I think that we have become quite familiar with the term precision medicine which in the case of cancer is to do with which specific things have gone wrong with the cell at the genetic level and targeting those problems versus broad brush.
Let's-stop-things-from-growing-fast adaptive therapy is an emerging area which is much more about, “let's take things that are not only complex but also fast moving and evolving and let's adapt whatever it is we're doing,” whether it's one drug or 20 drugs at the same time, and let's move with the disease.
Intuitively, we all know that that's going to be the future. Practically, it's not that easy to implement. But that's the current sort of emerging definition is that—especially in cancer, which is not a simple slow-changing disease, it's an extremely complex fast evolving disease—that you don't only treat the heterogeneity that's sort of the mixed-up-ness of it with a bunch of things that are targeted to you as an individual and your individual problem, but you're also moving those things and changing those things in real time as the cancer is evolving.
Juliet: Right. So this is where the analogy of adaptive design and clinical trials may be helpful. We've read that some statisticians and physicians have written that that sort of adaptive design for clinical trials specifically can be suboptimal or even harmful to the accuracy of the trial. And so I'm curious…what are the challenges of using adaptive therapy in oncology? What could go wrong?
Dr. Jack: It's enormously challenging. There have been some brilliant…we have cut out the wood from the trees and separated the wheat from the chaff and whatever other lovely metaphor you wish to use in medicine as a result of what we now live and die by, really, which is what we call evidence-based medicine. The crux of which in medical oncology is the randomized controlled placebo trial, which we have used now for a long time in order to practice good medical science.
There are, however, some pretty interesting results that we've gotten from that, which is the fact that we still have made some appalling progress really only in a lot of cancers as a result of that methodology—A versus B—which is fundamentally what we're talking about. Which is better, A versus B. That randomized, and see what comes out statistically, is relatively easy to do, but it turns out to be quite expensive for a lot of cancers. It has not moved the ball down the field very far and I'll come onto why that it. But that process has stopped us from poisoning people.
And so with quackery…I just came back from the Science Museum the other night and learned all about how the son of the Tsar of Russia, the last Tsar of Russia Nicholas the Second, who had hemophilia—the royal disease—was being treated with all sorts of mystical and magical stuff, and I just looked at the medical cabinet—of which was probably the most advanced medical cabinet in the world—and this is only 100 years ago. I'm pretty glad we're not there now.
We’ve gotten rid of that problem, but where we're at is that we've kind of made a lot of headway in simple diseases where there's one thing that you can attack with one agent or molecule technique and learned over time which is the next best thing, and so forth. But complex diseases, the ones which we’re kind of left with now, are the diseases of ageing, especially cancer, which is I think now probably number two in terms of the thing that's going to kill us probably will become number one as we get older and, you know, so-called healthier. We will have to face this problem today. I think it's already about a quarter to a third of people will die of cancer and half will die with cancer. We haven't made, in certain complex cancers, enough headway, and the result shamefully is that it is a disease that still kills half of the people that get it. This is overall, and there are certain cancers including glioblastoma, pancreatic cancer, advanced breast, bowel, liver, and so forth which have appalling success rates, and yet we spent billions and billions using the A vs. B methodology, which comes into my point about complexity. These diseases are complex. They're not a single problem. They are multiple problems that all feed off each other in order to create resilience, and they also get cleverer every time we try and tackle the problem.
They change the way they do things and evade the immune system more and protect themselves more and, you know, absorb more antioxidants so that radiotherapy doesn't work. But all these wonderful things, these jolly clever complex diseases, are not suited to the paradigm of A versus B because you attack one problem with one candidate drug and all the other problems get cleverer and form resilience and make the problem actually worse sometimes. So what adaptive approaches are doing is not saying that it's A versus B, but what are all the problems that you have? What is the molecular signature of your disease? Not like what does the histology just say it is on a piece of paper, but what does the gene sequencing machine spit out? And they end up being complex webs of interacting pathways that you can then treat with a combination of therapies, and then you retest every month, let's say, in some of the latest adaptive trials protocols. You might be re sampling the tumor either using blood-based biopsies or actually re biopsy the cancer and combining that with imaging and other biomarkers and so forth and then changing the therapy every month, not waiting five years, but waiting only a month before you change.
And this poses great problems for statisticians who are used to saying, “Yeah, but what's working in this? Is it A or B or C or D or E or F or G or whatever combination of things?” And ultimately what you're having to try to convince people of is that the A versus B paradigm has not worked. In the case of glioblastoma, we have increased life expectancy by the sum total of I think probably about 35 minutes in 40 years.
Well, it's a little bit more than that, but it's still utterly woeful the results, as a result of billions of dollars of A versus B trials, and yet it is largely because of the complexity and the fast-moving nature of the disease. And we need complex and fast-moving adaptive trials to be able to work out what it is that's going to work. And it's not so much whether it's A or B but it's the system. It's the diagnostics. It's the algorithms that help you choose. And then it's the monitoring that is fundamentally what we're testing. You're no longer testing A vs. B.
Then comes the bigger problem, which is the drug companies that are trying to sell A or B, or want to get paid for A or B.
And it's incredibly hard when you've got six drug companies providing a sixth of the solution each. And so that's the problem we're getting. It's a statistical issue but it's also a fundamental health care industry issue. How do you reimburse for what if you're changing things the whole time?
Juliet: So, you have pharma companies who are influencing the decisions that they're making; you have statisticians who have never really dealt with problems in this way; lack of standardization in labs; lack of familiarity with certain tests; it's very expensive… I believe that there are tons of obstacles that need to be dealt with. Let's assume for a moment that you could mitigate all of them. This still feels very specific to each patient. How can adaptive therapies become more accessible? Do physicians need to be educated about this, or do patients need to take more ownership over their own treatment as it's changing constantly?
Dr. Jack: Yes. There is a magic wand which I'd love to have and wish the world was 50 years more advanced than it is, but it isn't.
And we have to get to these through these steps, incrementally, and hopefully as fast as possible. I think the best example that I've seen of how to topple down some of these barriers and ultimately get to universal access to this kind of approach is the agile brain cancer mission. It was only about a year ago, the 28th of January 2018, that Baroness Tessa Jowell gave her final speech in the House of Lords, which got an unprecedented—and I think unique—standing ovation, where she explained about her illness, which was glioblastoma, probably the least treatable and most difficult cancer that we really know of.
And she was explaining that we have to change the way we work in medicine to be more collaborative so that we're not isolated in terms of being the one consulting with one patient, but actually accept the views of people, learning stuff from all over the world. That was one element of what she was asking for.
And her other call to action was to bring about adaptive combinations of therapies and to make them accessible so that they wouldn't remain in the realm of the very well networked and well off, but was something that became the standard of care.
And she actually put it very beautifully. She says you know that there are challenges, but let those challenges not put this into the “too difficult” box.
And what I have seen over the last twelve months, and being given the great honor of being one of the leads in the brain cancer mission, is the coming together of four very important initiatives, one of which is to better understand using more multidisciplinary science the fundamental biology of what makes brain cells do what they do when they become glioblastoma. Fantastic, we’re no longer just cancer researchers, we're getting astrophysicists involved and all sorts of wonderful embryologists and really big multidisciplinary sets of people, so fundamentally getting that bit cracked was one part.
Perhaps more relevant to your question is then how do we improve training? So it's all well and good to have wonderful new algorithms and combinations of drugs. But if the clinicians are scared by that then you need to train them to be less scared.
Let me ask you a question. How many neuro oncologists, that means cancer doctors that specialize in the brain, do we have in the UK?
Juliet: I have absolutely no idea. 100?
Dr. Jack: None.
Juliet: Oh, wow.
Dr. Jack: We do not have a single dedicated neuro oncologist in the country. So we have breast cancer doctors that also do. So not only do we not have people that understand the molecular basis of the disease and how to more frequently adapt our standards of care in more real time, but we don't even have people that specialize in that particular area. So we've got a long way to go, and we have set up a new training scheme which will is Tessa Jowell neuro oncology training schemes to all colleges that want to specialize in that. And that's a very, very important component.
The third component is patient centricity and patient-centered design, whereby everything that we want to take forward in the future is going to be designed for and with patients as opposed to just expecting them to enjoy whatever benefits of the science and new drugs are going to be offered them. What do you want out of all of this spend and out of all of these efforts? What do you want out of your disease and what do you want to try? What do you want out of research?
Which leads to the last track, which is the next generation of treatments and trials for the brain cancer mission. And that brings together Cancer Research UK leading academic institutes. It's being led by a brilliant man called Professor Colin Botts out of Birmingham. He's a neurosurgeon and the brain matrix is the first brain cancer adaptive combination therapy platform that has national scale ambition actually anywhere in the world.
And I think that if you're going to take this out of the realm of ultra-exclusive, bespoke, personalized medicine, where only a few of the most connected and well-off people in the world get to enjoy better outcomes, and you want this to become universally accessible—but also this is important again for people investing in this, and meaningful to industries so that they want to invest in this and put their money behind it—it's got to be with a goal towards universal access.
And if that is possible, it's only going to be possible as a result of addressing those four things. Basic science. Training. Patient centricity. And the implementation of Next Generation adaptive trials, which then become next generation adaptive treatment. While technology is perhaps in each of these areas, you are spending less time thinking and are making adaptive therapy and oncology easier to scale on your smartphone.
Well look, none of this would be possible without connectivity. There's connectivity at many levels. There is absolutely no way that you can adapt someone's therapy in more real time without knowing what it is doing, what it is that's going on with them.
And if God forbid I get something really ghastly, I want the next Jack who gets something ghastly to benefit in more real time from what's going on with me.
So, connectivity and all technologies which enable information sharing and information processing is an absolute backbone. Without the web, without interconnectivity between patient records, without the ability to also process those data—in brackets: safely, securely and with everyone's permission…close brackets—you don't have anything. You then go back into the same cycles of, let's control the experiment. Let's wait for a few years, let's then publish it and talk about it, and then let's change guidelines at the pace of a geological epoch. Well no, not quite as slow as tha, but certainly it takes the long cycles to make any real change. So, connectivity is a huge thing.
I think obviously the price of genomics has come down a lot. It was only in 2007 I think that the science paper came out with a publication of the first full human genome and I think it was 10 countries, three billion years, and a billion dollars to produce it now costs about a thousand bucks to get your tumor sequenced.
That's quite extraordinary. And so obviously the computational technologies, but also just the hardware and the global infrastructure for being able to do next gen sequencing, is absolutely transformative for our understanding of cancer. But in my view, that's because we've been treating the problem too simply. And so that's a part of the technology stack that is required.
And then I think an understated part of this is the role that the patient plays in helping to use technology themselves to help the community, both of patients but also of science, to know what's going on out there.
And also you know simple things like, OK, I've used tech and connected information and connected patients to be able to know what's going on and be more informed and then reconsidered ever more sophisticated, changing, possibly adaptive therapies, and then share what is going on as real world evidence to become part of the backbone of the scientific literature on what works and what doesn't. And again, without that, this wouldn't be possible.
So, ten years ago, everyone was talking about how we should get combinations of adaptive therapies. And I said “Yeah, but it cost billions,” and in ten years it's turned into something that really is less than the cost of lots of the other parts of the equation, such as imaging for insights.
I mean, if you think about where the billions are today, I can't help but think again about pharma companies. How do they adjust their operating model to work in the adaptive therapy space?
First of all, I think that we should not try and touch the big five cancers where we have seen incredible progress. And we've put more money towards understanding those simple things that we can do to really improve outcomes. You know you have to let the breast cancer…to see that there are things that you can do to early detect, to prevent, and to treat that are highly effective.
And so, to concentrate on the things which you can prevent and are incredibly hard to treat which have traditionally been less interesting to pharma—as I say I think about 50 percent, 40 to 50 percent of patients today in in the world with cancer, which is about 15 million new cases each year—have got intractable cancers. Of the 30 million people that get cancer each year, I actually think those numbers are a little bit higher than what people project. I'm also including countries that don't report as well. But it's a huge thing. I mean, we're talking about 15 million people a year have got cancer. There is no effective treatment. By that I mean probably anything you throw at the patient will give them maybe a 10 percent response rate. That's a lot of people and that's interesting to industry.
I mean, being very brutal about it, it's not only interesting to the mothers and fathers and brothers and sisters and the children of parents that have got these cancers, but it is interesting to industry. So, let's choose the cancers that currently we've not made much headway with that constitute part of this half of all cancers and let's work on that.
They also happen to be the cancers that are probably most suited to combinations of evolving therapies: adaptive oncology, as we call it. And so, it makes a lot of sense. It's a new market. It's untapped. It has huge unmet needs, which technology can now help us address.
And I would say the main thing is for industry to accept that there is this unmet need and to put some effort behind it, which does not disturb the current business model.
Juliet: One last question that naturally flows from that. What advice do you have for founders who want to build businesses in adaptive therapy, who see the opportunity in industry and want to take advantage of it?
Dr. Jack: The biggest advice is to make sure that you leverage this whole movement towards real world evidence, so anything that is collecting real world evidence and helping build that as part of the industry's need to understand actually what is going on with the patient, not in terms of imaging and overall survival but really, how does it affect the person's return to work? And how they're feeling? And how they're able to look after themselves independently? That's one thing which I would really encourage startups to do, is to be absolutely 100 percent on the side of the patient, because the patients want to report the record and predict what's going to go on in their homes and in their bodies in the real world and be able to feed that back into trials processes and as companion diagnostics for drug brands and for oncology brands. That means the pharmaceutical industry essentially. So that's one thing.
I think people that are working on ways of taking very complex tumor biology data, be it genetic, epigenetic matter, or immunological data, and being able to parse that into things that do not create ridiculous amounts of work for oncologists and their multidisciplinary teams, but rather digests the ends of things where you can simplify and take action. I think it is a huge thing that involves data analytics and visualization, which is a real kind of visual, graphical, and potentially also transformative way of turning diagnostics into something that's super scary and too complicated into things that will actually make a difference to people.
And I think the third thing is that people who are much more in the deep tech end of things, who actually may be thinking about discovering new drugs, should not think about single blockbusters. They should think about multiple pathways and they should think about how to build predictive models, and not of of how to treat disease at one point in time. But what do you need? What points do you need to hit in order to treat a disease and then how are those points going to change? You are no longer a drug company thinking about a target, but you're thinking about becoming a drug discovery company that looks at the interacting targets and how they're likely to change the course of the disease as it evolves.
I think the fourth one, which I've personally got an interest in but I think is quite heavy for startups, because it involves hardware and imaging and stuff, is the whole area of hyper polarized magnetic resonance imaging. I have a very deep interest in this as there's an international infrastructure of MRI scanners, but we can quite interestingly and easily enhance the resolution of that existing infrastructure by something like five to six orders of magnitude—that's ten thousand to one hundred thousand times—using very clever, nontoxic agents like hyper polarized pyrovate or glucose, which you can squeeze into your arm and it will light up your cancer like never before, to a resolution that people think is quite unheard of. But it's real! And I think that is going to transform the way that we monitor cancer. We won't have to wait three months to know whether things are working. We should theoretically, and I hope the right startup comes along and does this, be able to know the next day after you've had your therapy whether it's working or not.
Juliet: Well, Dr. Jack, thank you. I certainly hope the same. It is hugely helpful to hear you share some of your perspectives and experiences, and to understand how we can better evaluate technologies that are being built to make adaptive therapy the new standard of care. So, until next time on the Mosaic Ventures podcast, thank you for joining us.
Dr. Jack: Thank you so much.