This transcript has been edited for readability:

Government Regulation: The FDA’s Take

McGinley:       Thank you all for joining us.  I’m Laurie McGinley.  I am a health and medical reporter here at The Post.  I’m delighted today to introduce to you Scott Gottlieb, the commissioner of the Food and Drug Administration.  Since being sworn in in May, Dr. Gottlieb has been very, very busy.  He’s been working on the opioid crisis, generic drug approvals, and the high cost of drugs and drug development, which is something that honestly, we don’t often hear about from the FDA.  So, that’s very interesting.

On cancer, Dr. Gottlieb has personal experience.  He was successfully treated for Hodgkin’s lymphoma in 2007.  And that gives him a unique perspective along with other cancer survivors.  One final note before we get going.  For those of you in the audience, and people watching online, you can send us your questions on Twitter, using the hashtag PostLive.  Or you can comment on our Facebook livestream.  Dr. Gottlieb, thanks again for joining us.

Gottlieb:          Thanks for having me.

McGinley:       So, I want to start by talking about something that we’ve already heard about today, which was the big news last month, when the FDA approved the first of its kind immunotherapy to treat childhood leukemia.  As I’m sure most of you know, this involved an approach called CAR T cell therapy, that uses a patient’s own genetically modified cells to seek out and attack cancer.  At the time of the approval, you said that it was a big deal, and that we were entering a new frontier.  Why was this advance so important?  Why was the approval so important? And what were you referring to when you said that there is more to come?

Gottlieb:          Well, it was a significant milestone on multiple levels.  I think the most meaningful of which is that it was the first gene therapy product approved in the United States, and there had been one previously approved in Europe, but this was the first approval in the United States, and it was the first approval of a product that looks to be highly effective.

And I think we’ve been working with gene therapy as a concept for many years—for decades—and we’ve had a lot of false starts and some tragedies along the way, and a lot of challenges.  And I think we’ve finally reached an inflection point in science where we have the techniques sufficiently perfected to actually allow these products to be translated into therapies for humans, and that’s what we’re seeing.  And this was the first one, but there’s literally hundreds behind it.  I think we have more than 500 active INDs in-house at FDA for various gene therapy products.

And so, we’re going to start seeing hopefully other approvals of products, and the inflection point, I think, in the science of gene therapy, probably came with the advent of vectors that were effective at delivering the gene and not causing side effects.  For a while, we had—we were using vectors that either wouldn’t effectively deliver the gene or themselves would have certain side effects associated with them that created toxicities, and I think we’re finally at a point in science where we understand enough about how to do this; that we’re starting to see the use of more effective vectors, and in some cases, highly effective vectors.

And that really was the turning point in the availability of these treatments.  The other significant component of this was the application itself.  You’re effectively using the tools and gene therapy to modify t-cells to allow the patient’s own immune cells to attack their cancer, and that’s obviously a highly novel approach to the treatment of cancer, where we’re trying to use immunology, and immunotherapy to personalize a treatment to attack someone’s cancer.  So, the days of systemic therapy where we poison the whole body with the hopes that we would poison the cancer cells more, a lot of cancers are still treated that way, but we’re now starting to see immunotherapy become more an integral part of cancer treatment. And the use of CAR T is another big milestone, turning point, I think, in that approach.

McGinley:       You said at the time that the FDA stands ready and has taken steps to expedite the review and approval of these products, what were you referring to?

Gottlieb:          Well, so, we’ve obviously—we have certain authorities like breakthrough therapy pathway, which allow us to expedite the approval of products that target unmet medical needs, and look like they’re going to be a significant breakthrough, a significant event in the treatment of a disease that otherwise isn’t impacted by available therapies.  And so, we have things like fast track and accelerated approval that you’re familiar with, where you can get earlier approval, for example, on the basis of a surrogate measure of clinical benefit.

But the other thing we’re doing when it comes to some of these new biologics, of which CAR T is one when you look in the FDA’s biologic center, at the products like the gene therapy products, or cell therapy products, is we’re starting to take steps to have more engagement with sponsors very early in the development process, because we find with some of these very novel platforms, there’s more uncertainty about what the expectations are, especially when you’re the first one to come through.  And so, for a lot of these very novel technologies, we’re having much more extensive earlier engagement, even pre-IND, even before companies come in, ask for permission to submit an application to FDA to ask for the ability to go into human studies, we’ll take meetings with them to discuss the potential development approach.

McGinley:       I think that for most patients—for most cancer patients, although I think this probably goes for any disease, of course—that what patients really care about is safety, effectiveness, and increasingly about cost.  And I’d be remiss if I didn’t say that the CAR T cell therapy clocks in at $475,000 for a treatment, which is intended to be curative.  But we’ll get back to cost in a second.  I wanted to start with safety for a minute.  So, in talking about your intention to expedite things at the FDA and makes things more efficient, and certainly President Trump talked about that as well, how much faster can the FDA get?  The FDA is already one of the fastest drug review agencies in the world, and do you have any concerns that at some point, if you keep accelerating, that you’re going to compromise safety?

Gottlieb:          Well, I think when we talk about efficiency, and you talk about speed in the review process, when I talk about it, I’m not talking about review times.  I’m not talking about application review times, because you’re right, we are very efficient.  We review applications in a matter of months—eight to ten months, depending on the review cycle that an application is on.  And most of the applications for the new molecular entities are—a decision is made on the first cycle of review, so they’re not undergoing multiple cycles or review by the agency, where the agency reviews it, says, “We can’t make a decision,” asks the sponsor to do more work.

So, it’s actually—it’s very efficient.  I think where we can have an impact is on a development timeline, where the FDA does have an impact on how long it takes to actually do the clinical studies to develop a product by virtue of the kinds of requirements that the FDA asks for.  How you structure clinical trials; the kinds of clinical evidence that the agency asks for; how you do pre-clinical studies, maybe in toxicology studies, whether you’re doing them in assays or in animals, for example.  There’s different ways that we can incorporate new technology and new methodologies to make that process more efficient.

And I think if we’re doing our job in terms of thinking about the new science, and how you can make the process of drug development itself more efficient, and more scientifically rigorous, this isn’t a binary assumption; this isn’t a binary choice where if we can reduce development timelines on average, we’re going to have less safe drugs.  I think you can actually reduce development timelines, and have a higher assurance of safety and efficacy of what you’re doing is scientifically sound and you’re incorporating better methodologies.  And that’s what we’ve talked about, and I talked about that recently in a series of speeches I gave, looking at things like seamless clinical trial designs, or master protocols where you allow the study of multiple indications within the confines of the same clinical trial, or you look at enriched trial design, where you’re successively enriching the clinical trial for characteristics that you think correlate with patient response.

And so, by the time you’re done with the clinical trial, you have a whole set of descriptive data that describes which patients are more likely to have a therapeutic benefit, but you’ve also potentially shortened your development timeline, because you’ve enriched the trial for patients who are more likely to benefit, so you can see the benefit earlier.  These are just some of the approaches we’ve found opportunities to use assays—cellular and tissue assays that could be more descriptive in terms of looking at toxicology, and then maybe doing studies in animals, and we’ve incorporated those into development process.  So, these are places where I think you can adopt better science, and actually have a more efficient development process.

The problem is that we’re reluctant to change sometimes, and if we don’t change, sponsors don’t change.  So, people—if people have done something a certain way for a long period of time, and you have comfort with the certainty of the results you’re going to get, it’s hard to change, especially with so much at stake.  If what’s at stake is the ability to tell whether a drug is safe and effective, you don’t want to change methodologies until you’re very certain that you’re actually going to get a better result.  And so, innovating that process, from both a regulatory standpoint as well as on the part of sponsors, is hard, it’s slow, and I think it needs a concerted effort by leadership across the entire community, because people—you don’t want to make a mistake.  You don’t want to switch from one methodology to another, only to find out years later that actually the first way—your first approach was the better way of answering the question.

McGinley:       Well, and this would be why the drug companies would need some additional contact or reassurance from the FDA, that they’re going down the right path.

Gottlieb:          This is why we need to do things—this is why things are done in a transparent process, often through a notice and comment, either rule-making or guidance, but in order, I think, to give the development community comfort that a new methodology can not only be a better approach to something that you’ve done, but actually will be embraced by the regulatory authorities.  We have to go through a very deliberate process where we have workshops, where we socialize it, and where we ultimately develop written guidance documents articulating what the new standard would be.

And that’s what we’ve tried to do, so we’ve put out a series of documents related to guidance on how to do clinical trials, how to use Bayesian tools, statistical tools in clinical trials, or adaptive approaches.  All of these things that I’ve talked about, or incorporated different kinds of objective endpoints in the clinical trials as a way to evaluate both safety and benefit.

McGinley:       There’s a lot of buzz about real-world evidence.  This is a term that kind of eludes me.  Can you just explain what does that mean? Does that mean using information from electronic health records to support applications?  Or can you ever see a situation in which that kind of information would supplant a standard clinical trial?  Or what exactly are we talking about here?

Gottlieb:          Well, typically when we talk about real-world evidence and the definition—it means different things to different people.  But typically we’re talking about practical, clinical data that’s gleaned from routine clinically care, routine clinical practice, but it could be something that’s—it could be data that’s generated in the confines of a more structured approach, like a prospective registry, and so often times, especially on the medical device side of our house, a lot of the data that is used, even for purposes of regulatory decision making is derived from prospective registries that are rigorously done, but it’s data that’s collected on top of routine clinical practice.

And so, it might be data that’s collected in the context of an electronic health record.  We do have wide latitude as a regulatory authority to use real-world evidence to supplant our clinical trial requirements, and so we do have latitude, for example, to approve a drug on the basis of a single study, and then supplant that single study with confirmatory evidence—substantial clinical experience.  And that could be from—derived from real-world evidence.  And so, this already goes on; it especially is used in the context of drug safety to evaluate drug safety in the post market, but it’s also increasingly used to help augment what we know about products on the efficacy—on questions of efficacy, as well, particularly on the medical device side, but also on the drug side of our house.

McGinley:       So, your goal in all of this is to get more drugs on the market, stimulate competition, and hopefully that could curb or maybe even drive down drug costs.  And you’ve also said—well, first of all, is that correct?

Gottlieb:          Well, I think my goal running the agency is to get effective therapy to patients, and get safe and effective products to patients, especially when we’re looking at areas of new technology, where some of these products have the potential to be transformative in human health.  And I think that we want to make sure that we have an efficient route to market.  The CAR T is one example of that, that this has the potential to transform the treatment of a cancer.  The verdict is still out; we obviously have very good evidence about the short-term benefits.

We’re going to need more evidence around the long-term benefits and the sustainability of the response, but these have the potential to be transformative treatments.  I mean, we do have the potential with some of these new applications, like gene therapy, to cure inherited diseases, to cure things like beta thalassemia, or blood illnesses like sickle cell disease.  It isn’t out of our reach within the next decade that we might be able to dramatically alter the course of some of these diseases.  So, my goal is to make sure we have a regulatory process that’s as modern as the science we’re being asked to evaluate and get to the patient sufficiently.

McGinley:       Well, you have said—

Gottlieb:          And to try to bring down drug costs.

McGinley:       I was going to say—because you have said that high drug costs or prices, depending on how you want to put it, are a public health issue, which I think is completely legitimate, considering the studies that show that some people are not taking their cancer medications because they can’t afford it, or people are facing high co-pays.  Well, last January, President Trump said that the drug companies were getting away with murder on their pricing, and then there was a big flurry of activity about an executive order on drug prices, possibly coming out.  And then it just seemed to all disappear some months ago.  Where does that stand?  Where does the executive order on drug pricing stand?  Does it exist?

Gottlieb:          Well, I’m not—I don’t speak for the White House.  You know that, Laurie.  Good try.  I could talk to you about what we’re doing.  I think what we’re doing with respect to drug pricing and drug competition could become a component of any policy that the administration puts out, but I don’t know exactly what the White House is working on.  You’d have to direct those questions there, but obviously, we’re trying to do things on the generic drug side.

We also have to think about the new drug side.  I think that we’ve seen examples where when the second and third new drug enters a category, you see substantial price competition, and one of my concerns is that it’s becoming, because of development costs becoming so high, and the amount of investment you have to put in is so enormous now, you’re seeing fewer and fewer categories where companies want to be the third or fourth to market.  In fact, you’re seeing oftentimes after the second or third drug comes to market in some of these categories, other companies will pull out.

And so, that could be a challenge going forward, because you might see less competition in these new drug categories.  And to just pick up on the issues around the CAR T, that’s a particular challenge when you have therapy where it’s going to be very highly specialized, so this treatment is targeted to fewer than 200 patients a year, although the indication might be expanded, but also the cost of goods, the cost of the product itself are not trivial.  In this case, it’s probably tens of thousands of dollars—the product itself.  And, so that’s going to be a particular challenge, how we finance this, and how we make sure patients can get access to it.

McGinley:       And it’s a one-time treatment.

Gottlieb:          And it’s a one-time treatment, and so you have a one-time administration, a one-time cost.  And so, how do you—these sort of policy questions that are outside my current purview, but how do you think about how you amortize those costs, spread them out so it’s not a one-time hit to the system, especially if we come up with the ability to cure a pediatric disease, an inherited disease, where you might want to deliver it to everyone at one time?  Those could be substantial costs to the entire healthcare system.

McGinley:       What did you think you were—I’m not sure how much you were involved in the Novartis—which is the maker of the CAR T cell therapy called Kymriah—decision to issue this money-back guarantee, that if the patient did not respond within the first month, that the patient wouldn’t have to pay for the therapy.  What do you think about that?  And are we going to see more of that?  And what happens when a third of those patients relapse after a year?  Should they get a partial refund, or what should happen then?

Gottlieb:          Well, these are—I think we’re going to see—to answer your first question, I think we’re going to see more outcomes based pricing for a lot of these treatments, especially treatments where the response is very binary.  You can get a dramatic result or no result at all.  In terms of how we make a decision about what endpoint you’re measuring, those are the kinds of things that are going to be negotiated out between the technology developers and the payers, but that’s also the complexity of having these outcomes based contracts.  Sometimes it’s hard to decide what benefit you’re going to measure, or how long you’re going to measure it over, and that’s been the challenge with these in the past, is often times, drug response isn’t binary or it’s hard to measure.

But I think you do have treatments now—this one and others, perhaps, behind it—where you will see a binary response.  You might not know the durability of the response at the outset, but you’re going to see a clear effect, and you’re going to know whether or not the patient had a clear response to a product.  And so, those kinds of constructs will lend themselves to these outcomes based contracts.  That’s why I think we’re going to see more of it in the future.

McGinley:       I was interested—you said in your statement last week about trying to lower drug development costs, that you hoped and expected that the savings—some of the savings would go to consumers.  What makes you think that that would happen?  And is there anything that you could do to make that happen?

Gottlieb:          Right, look.  It’s not a one on one correlation.  There’s complex economic models and I know people in a competitive market and a dynamic market price things to the price that they can derive and what they perceive the value to be, and what the value is to the end recipient.  So, simply lowering the development cost, products aren’t priced based on some multiple of what they cost to develop them.  But from my standpoint, when you look at development costs that can clearly top a billion dollars in direct outlays, and I know there’s estimates about what the time cost and capital is, and what the imputed cost of all the failed clinical studies are, and then that gets you to figures like $2.6 billion, which is a figure that’s sometimes quoted by the industry.

But we can all agree, it’s really expensive.  And I think if it continues to remain very expensive, and continues to grow at the rate that it is, and a rate that, by many measures, is in excess of even the significant rates or increases in the prices, we’re going to see fewer drugs come into these categories, and less competition.  And that certainly will impact prices, because we know, and we have objective evidence—I don’t want to cite specific drug categories but we do have objective evidence where when the second or third drug comes to the market in a new drug category, you see significant competition.  You see payers have the ability to negotiate significant price discounts, often times in excess of even what the discounts are that are extracted by government in Europe.

And Hepatitis C is one such category, to name one example, where you saw pricing come down significantly according to public reports.  I think you’ve published after multiple drugs came into the market, and so you want to have an economic model that allows the second and third and perhaps the fourth drug to come into the market, not only because it brings competition, but it also brings product differentiation that could be important to a subset of patients that might only respond to one particular iteration of a product.

And I think when you’re looking at direct development costs that easily top a billion dollars for some of these products, the economic model becomes more and more challenging, looking out into the future, and that does concern me.

McGinley:       What about biosimilars?  How big an answer are they to the problem?  The FDA just approved its first biosimilar for a cancer drug for Avastin.  And I think the agency, correct me if I’m wrong, has approved maybe a half a dozen other biosimilars for other diseases.  What are you seeing in terms of the adoption—the uptake by the doctors, by the providers, and are you comfortable with the rate of adoption?

Gottlieb:          Well, we currently have ten applications in-house.  We have 27 sponsors that have asked for guidance on applications, and so the pipeline in-house at FDA isn’t exceedingly robust, but behind that, we see a lot of development activity.  And so, I think we’re going to see a real pickup in the rate of biosimilar development.  I think we’re at the early stages of biosimilar, similar to where we were 30 years ago with generic drugs, where adoption is slow; there is reluctant on the part of providers to switch over, certainly to switch patients off of therapy that they’re on, onto a biosimilar, or even to embrace a biosimilar, especially if you’re thinking about curative therapy.

And so, we’re going to have to continue to educate providers.  We’re spending some money on the part of FDA to go out and do some public education campaigns to educate providers about the robustness of the process that biosimilars go through, but I think it’s going to be slow adoption.  I think this is an opportunity where the payers might have a role to play by guaranteeing perhaps a market share to some of the biosimilars that are coming onto the market, and then driving the adoption themselves, because what I worry about, and I’m taking off my FDA hat for a moment, but what I worry about is if the adoption rates continue to be slow, then the potential manufacturers of biosimilars won’t see this as a viable opportunity, and won’t make the investments in the first place.

If they don’t think that they can capture 20- or 30% market share within the first five years of being on the market or whatever the economic model is, they might say, “You know what?  This is a category we’re going to stay away from.”  So, I think the payers have an opportunity to perhaps think about guaranteeing some market share, and then putting the obligation on themselves for having to drive adoption. If we don’t think a little bit differently about this, and do something potentially disruptive, it could be a very slow ramp, and I do think that these have the potential to provide important competition in categories where the products, right now, could be exceedingly expensive.

McGinley:       You’ve said very nice things about Richard Pazdur, who’s head of the Oncology Center for Excellence, and I think has instituted some of the things that you’re talking about in terms of new types of trials, and I was interested to find out—my understanding is that the funding is kind of huge up on the oncology center, and that the funding that was—that people thought was going to get there has not gotten there.  Can you tell me what’s going on there, and how it might be resolved?

Gottlieb:          Well, to your point about Rick talking about some of the things I’m talking about, Rick actually pioneered some of the things I’m talking about, and wrote my talk points for me.  So, I’m talking about things that Rick really invented and championed, and I have adopted them as a matter of policy.  So, the issue with the Oncology Center for Excellence is this is a combined program between our biologic center and our drug center, to try to consolidate the different aspects of oncology product development into one program, and this is a format that we would like to replicate in other therapeutic areas.  This was the first one that we did.

The money that was supposed to be allocated to the development of this program was allocated through the NIH, and it was supposed to be given to FDA and not to get sort of too government speak on you, but it’s not so easy for the NIH to write us a check.  We didn’t know that at the time it was allocated in this way, so we’re going to need to see how we could fix that in order to make sure the program gets the allocation that Congress intended.

McGinley:       I think since this is a cancer event, I would be remiss if I didn’t ask you about your tobacco initiative.  And during the summer, you announced a new comprehensive tobacco regulatory framework with a couple of different prongs, so to speak.  Could you tell us your thinking about it, and what the next step is on that?

Gottlieb:          So, this is an effort to put nicotine at the center of our tobacco regulatory efforts, and see—view nicotine really as the problem, both the problem as the potential solution to cigarette addiction.  What kills people from cigarette smoking isn’t the nicotine, although nicotine isn’t a completely benign substance, especially in a developing brain of a young child.  What kills people are all the carcinogens that are attached to tobacco when tobacco is combusted.  And so, the first part of this effort is to regulate nicotine levels in combustible cigarettes, to try to reduce nicotine levels to minimally addictive or non-addictive levels, and reduce the addictive attractiveness of cigarettes, so we don’t hook a whole new generation of smokers onto cigarettes, and so that people who want to get access to nicotine look to other, potentially less harmful products than the tobacco, than the combustible cigarettes.

At the same time, we pushed off application deadlines that were due on certain non-combustible products including e-cigarettes.  If we had not done that, they were facing a cliff where they would have been swept from the market.  And so, what we’re doing is putting in place regulations to put those products through an appropriate series of regulatory gates while we allow them to continue to advance and innovate.  This is new product innovation that has a potential to be a lot less harmful than combusting tobacco for those people who still want to get access to satisfying levels of nicotine.

So, our view was at the very time that we’re taking decisive action to try to reduce nicotine in cigarettes to non-addictive levels, we need to provide an alternative for adults who still want to get access to nicotine, and be new—what we call ENDS; electronic nicotine delivery systems, of which e-cigarettes are a prominent component. Provide one alternative, potentially, for a safer route for patients to still get access, or people to still get access to nicotine.  At the same time, we’re going to be taking steps also to look at the medicinal side of our house, and how we can allow more product innovation when it comes to nicotine replacement products regulated as drugs, so the OTC products.

Again, to make sure that we’re providing proper avenues for people who want to get access to nicotine, adults who want to get access to nicotine, to be able to do it without having to combust tobacco.

McGinley:       What do you say to people who say, “That’s all fine and good, but the FDA could have done the nicotine—reduced nicotine years ago, because it had the authority, and it never did.”  And in the meantime, now the regulation of these e-cigarettes, which some—which people don’t really know what’s in them, has been pushed off a number of years.

Gottlieb:          Well, the regulation of the e-cigarettes wasn’t pushed off.  The newly—the so-called newly deemed products of which e-cigarettes are one are fully regulated by FDA.  So, we can enforce GMP standards, we can and will continue to do inspections of manufacturers; we require warnings on labels; all of those regulations are still in effect.  What we’ve pushed off are the product applications, and under the provision the products would have effectively have been swept from the market.

And so, I don’t think it would have been good public policy at the very point at which we’re trying to regulate nicotine to non-addictive levels in the combustible forms of tobacco, not to provide a pathway towards alternative delivery of nicotine, for people who still want to get access to it.  And we’re going to have a whole generation of smokers, some of whom are still going to want to get access to nicotine through vehicles that could replicate the experience of smoking, and what are we going to do for those people?  So, these provide a potential solution, a potential alternative for adults who want to get access to them.

So, we’re going to use—the reason why we picked the amount of time we did for how long we were pushing off those application deadlines is because that’s the amount of time it’s going to take us to get implementing regulations in place, to put those products through an appropriate series of regulatory gates.  But we’re still going to be continuing to regulate them and oversee those products, even while we extend the application deadlines.

McGinley:       Can you tell us before we wrap up a little bit about your experience dealing with cancer?  You were 33, and deputy commissioner at the FDA, and what happened?

Gottlieb:          I found the kind of lesion on myself that only a hypochondriac physician would have found.  It was very early stage.  It was stage 1A Hodgkin’s lymphoma.  It was just a single, almost—a mass that almost couldn’t be palpable, but I felt a fullness in my axilla—my left axilla, and ignored it like some people do.  Most people do for a little bit of time, and then one day, said, “This doesn’t feel right,” and went to the doctor, and was diagnosed with a CT scan.  And probably a CT scan that nobody would have had—nobody in this room would have had unless you were the deputy commissioner of the FDA, walking into a general practitioner’s office one morning.

I think that she had a heightened sense of concern that she shouldn’t miss something, but I’m grateful for that.  I think the—I went through a very traditional chemotherapy regimen; Adriamycin, bleomycin, gomblastin, dacarbazine, Rituxan, so it was a tough six months of chemo.  I did most of it while I was still working at the FDA, but I think the hardest part for me wasn’t the physical aspects of the treatment, but the uncertainty.  Probably the hardest parts were that first weekend when I knew I had a mass in my chest, but I didn’t know what it was.  I’m a physician, so I spent the whole weekend online and going through medical textbooks, and I said, “Could it be metastatic colon cancer?”  I was going through all the permutations of the worst thing it could be.

When I got the diagnosis, I was obviously—had some reassurance, because it was a very curable cancer, but there was still a lot of uncertainty with the treatment regimen, would I relapse, would it shrink?  And that was the hardest part.  I think the mental aspects of it were exceedingly difficult for me, and I think they are for most people.  The physical—

McGinley:       And how has it affected how you approach your job, or how you think about?

Gottlieb:          I think it’s affected how I think about my job in so far as for me, I was a patient who was told, “You have a 90% chance of a cure,” and I was trying to figure out how do I get 90% to 92% or 93%.  And the way to do that was to scour the literature and look for the most decisive studies I could find about how to structure my regimen.  I had to make decisions.  Do I use radiation or not?  Or do I try Rituxan, which was experimental at the time.  Now it is used more commonly.  So, I was looking for data.  I wanted data.  I think for patients who are diagnosed and told, you know, “You have a 20% chance of surviving five years,” that’s a very different situation, and what that patient is looking for is something new.  They’re looking for a silver bullet. They’re looking for something experimental, because for them, the available therapy has failed, and I think that there’s a lot of different kinds of cancer patients, but I think there’s two ends to that spectrum.

And I think about that a lot, because I was in a different position, and I think that when we talk about trying to get really rigorous data and very large clinical trials, that serves me—my kind of patient very well; someone who has available therapy that’s effective, but is looking to get their odds up.  But if we prolong that process in search of very pristine data, we might not be serving the other patient very well, the one who is looking for something early, something experimental, because for them, the available therapy just isn’t sufficient.

McGinley:       Well, Dr. Gottlieb, thank you so much.  Unfortunately, we’re out of time, but we’re really enjoyed having you, and we will be back in a few minutes, and thank you very much.

Gottlieb:          Thanks.

McGinley:       Thank you.

Killing Cancer: America’s First FDA-Approved Gene Therapy

McGinley:       We’re going to continue our discussion on some of the new frontiers in medicine today.  And before we get started, I’d like to introduce the speakers who are on stage with me.  To my immediate left, we have Dr. Shannon Maude.  She’s an attending physician at Children’s Hospital of Philadelphia, also known as CHOP and is an expert in immunotherapy.  Next to her is Dr. Nirali Shah, an associate research physician in the Pediatric Oncology Branch at the National Cancer Institute.  She focuses on clinical trials with a specific focus also on immune-based therapies.

We’re also joined and I’m delighted to meet and to see Tom Whitehead, who is the co-founder of the Emily Whitehead Foundation.  Five years ago, his daughter, Emily, was the first child in the world to receive the groundbreaking immunotherapy that we have been talking about, the CAR T-cell therapy and that we will be talking about some more.  Her leukemia went into remission after being treated at CHOP and Emily will be joining us on the stage a little bit later.  So thank you all for joining us today.  So, Dr. Maude, I wonder if you could get us started and give us a little bit of background of how this clinical trial that Emily was treated in came to me.

Maude:            Well, immunotherapy has been something that’s been of interest to the cancer research field for many years; over 20 years and people have been studying ways to harness our immune system to be able to fight cancer.  It really wasn’t until recent years; I would say the last 10 years or so that that has been able to advance enough to the point that therapies were potentially effective and something that could move into clinical trials to test the safety and to test the efficacy.

So at Penn, Carl June had been working on immune-based therapies for cancer for many years and he, along with David Porter at Penn started a clinical trial in a different type of leukemia; chronic lymphocytic leukemia and treated three adults with a CD19 targeted immune therapy.

McGinley:       And CD19 is a protein that appears on the surface of?

Maude:            With B-cell cancers.  So there are different types of leukemias and lymphomas that arise from different cells in our immune system and one broad field are B-cell malignancies, leukemias, and lymphomas.  CD19 is a protein that is very widely expressed on those types of cancers and so something that is very attractive, to be able to target so that the immune system can find the cancer.  The principle behind this is that we take a part of the immune system, the T-cell in particular, which is very good at killing diseases, things that shouldn’t be in our body, and we teach it to be able to recognize the leukemia by recognizing that CD19.

McGinley:       And what does the HIV virus have to do with this?

Maude:            So the HIV virus is something that has been exploited to be able to reprogram and teach these T-cells how to recognize the leukemia.  So what we’re really doing is using a form of the HIV virus or something called a Lentivirus and changing it in a way so that it’s no longer HIV, but what we are doing is exploiting the ability of that virus to get into a T-cell and to deliver a piece of DNA.  So we’re really using it as a delivery vehicle to change the T-cells.

McGinley:       So as I understand it, Dr. June and Dr. Porter treated three adult patients.  They had good results.  They ran out of money so then they had to stop for a while and then when they received more funding, then they were working with CHOP to open the pediatric trial.  Is that the way it went, pretty much?

Maude:            Well, Dr. June and Dr. Grupp at CHOP had been collaborating for many years to study these therapies and there was definitely an interest to be able to move this into pediatrics and into one form of pediatric leukemia; acute lymphoblastic leukemia.  But in order to do that, we always have to first see that something can be safely given to adult sand that there’s some hint that it may work.  And so that was what came out of those first groundbreaking studies in the three adults and after those three adults were treated, then the first child was treated at CHOP.

McGinley:       So Mr. Whitehead; Tom, take us back to 2010 and as this was going on, the research was going on, you had a beautiful five-year-old daughter who was healthy and then she was not healthy, she was sick.  What happened?

Tom W:           Yeah, literally overnight.  It was Thursday, May 27th of 2010 and then she was healthy.  My wife had noticed that she had bruising on her legs and we look back now and she had some bleeding on her gums that week and I had noticed blood on the end of her nose but she felt normal until the day she was diagnosed.  So overnight, she had severe leg pains and by morning, we decided to take her to the pediatrician and Carrie [ph] took her; my wife Carrie and her mom, Pam, and I had to go to work, but by 9:00, they called me and said that Emily was in the emergency room and that there was really something wrong.  And over that weekend, which was Memorial Day weekend, they diagnosed her with leukemia.

McGinley:       So you went initially to Hershey Medical Center, is that right?

Tom W:           Correct.

McGinley:       And she had the standard treatment, which is, as I understand it, 26 months of chemo.  That’s a pretty tough regimen to start with.

Tom W:           Yeah, and she had a tough start but she did get into remission and she stayed in remission for 16 months; the first 16 months.  So in the beginning, we were told it was the garden variety kind of cancer and if you have to have a child with cancer, this is the one you want to have and it’s the most curable so we thought we could get through this and from the time Emily was diagnosed, I picked her up and looked her in the eye and said, “No matter what happens, you’re going to beat this because only the strongest kids in the world are picked to fight cancer.”  So it was pretty devastating 16 months in when we went for routine blood work and Dr. Jim Powell at Hershey—he was at Mount Nittany Medical Center but he’s local to us and he called me the next day and said, “Unfortunately, Emily’s relapsed and this almost never happens when chemotherapy is working for these kids.”

McGinley:       It works for 80 or 90% of children, right?

Tom W:           Yeah.

McGinley:       So she was, unfortunately, in the group it didn’t work for and then as I understand, you went back and forth a little bit between CHOP and Hershey and in any case, after a couple of back-and-forths, you ended up back at CHOP because almost amazingly, the T-cell trial was opening up earlier than expected and people were worried that Emily wasn’t going to be able to wait for the opening of the trial.

Tom W:           Correct, we had been told once about the T-cell trial but said Emily probably can’t survive long enough for the FDA to give us the approval to try it so you don’t wait for that.  We tried to get her to bone marrow transplant and we never met her donor, but the donor delayed us by three-to-four weeks and said they weren’t available to donate cells and halfway through that delay, she relapsed again and finally, Hershey docs came in and said, “Unfortunately, we don’t have any more weapons to fight her cancer.”  And when we paged Dr. Rheingold back at CHOP and said, “We’re coming no matter what because we’re not going home on hospice.”

And she said, “This is amazing to me because the last email I read yesterday was that we got FDA approval to try the T-cell trial.”

McGinley:       So at that point then, that’s the spring of 2012 and now you’re in Philadelphia and they extracted her T-cells and it took six weeks to genetically alter them to—

Tom W:           Yeah, that’s correct.  She was in isolation in her hospital room with no immune system for six weeks.  They told them, “We’re going to send them off to boot camp, then we’re going to train them to be in Army and when they come back in there, they’re going to go in there and kill your cancer.”

McGinley:       And what happened when she started getting the treatment?

Tom W:           So the first two days she did good and they explained to us and she was patient one in a phase one trial that you split up the dose; 10% the first day, 30% the second day, and then 60% the final third day.  The first two days she did good, but the second night around midnight she spiked a fever.  We went into the CHOP ER because we weren’t comfortable with not being in the hospital at that point.  As soon as we got there, she felt fine.  Her fever went away and kind of stayed up all night and she had popsicles and was doing good in the morning and Dr. Stephan Grupp came in and said, “She looks good and feels good and I think we need to get the rest of this dose in her.”  And we agreed.  We were 100% all for it.

But when she got the final dose though, she had so much cancer in her system and these cells revved up right away and there was so much cancer dying all at once that it overwhelmed her system and then she ended up in a coma on a ventilator, breathing for the next 14 days.

McGinley:       And what happened after that?

Tom W:           So during that time, while she was in the coma, the doctors were working around the clock to try to make her a success and give her every chance they could and they came up with the suggestion to try an arthritis drug on her to turn her symptoms around that had never been used on a cancer patient and that drug is tocilizumab and in her system, we found out later that she had over 3 pounds of tumor in her blood and that was all being killed at once and this IL-6 protein was out of control in her system, which was giving her really bad side effects and she was really swelled up to the point where they told us they didn’t think she could survive that night.

But when they gave her that medicine, things started to turn around with her blood pressure right away and I believe it was 14 days after the coma, she woke up on her seventh birthday.  But before that, Dr. Grupp came through the door one day and he said, “This is amazing to us and we’re cautiously optimistic that we see that Emily is improving faster than we expected for how sick she was and now, her b-cells are disappearing so now we think it’s working.  So again, she woke up on her seventh birthday, which was May 2nd of 2012 and then eight days later, which would be 23 days from her first dose of T-cells, they checked her bone marrow and she was completely cancer-free.

McGinley:       It’s quite an amazing story.  Dr. Shah, you do a lot of this work out at NCI and I’m wondering if you could talk a little bit more about the safety, these safety aspects of it and what was going on with Emily?  As I understand it, it was some protein called Interleukin 6 that was causing the problem and that Dr. June at Penn suggested the drug, which I can’t pronounce the name of and because his daughter has rheumatoid arthritis and she was on the drug and that it quieted down the immune system.

What have you seen?  You said something very interesting to me just a few minutes ago.  You said, “We have learned a lot from Emily Whitehead.”  What have you learned about from Emily Whitehead, both about safety and about other issues?

Shah:               So one of the main complications of giving CAR cell therapy is a side effect profile called cytokine release syndrome.

McGinley:       Which is what Emily had.

Shah:               Which is what Emily had and which is what they see with a lot of patients who are having a response to CAR therapy.  I’ll refer to it as “CRS”.  But CRS is essentially an inflammatory milieu that arises as the CAR cells go in.  The CAR cells go in, they see the leukemia, they expand and they start to grow in numbers and as that expansion is happening, there’s all of these inflammatory—we’ll say there are inflammatory cytokines that are then released in the body that are causing a lot of the side effects that Emily had.  They can lead to low blood pressure, difficult deep breathing, needing to go to the ICU.  Some could even lead to some forms of neurotoxicity, which you may have heard about as well.

So what we learned from Emily was that they checked her blood to see if there were any of these cytokines that were elevated.  IL-6 was amongst one of those cytokines.  Tocilizumab is an IL-16 receptor blockade, so it actually blocks the ability of IL-6 to affect your organs and by giving that blockade medication, the IL-6 is no longer causing the problems that it was causing in her body.  So as they learn about cytokine release syndrome, as they learn about the effectiveness of CAR therapy; what we learned from Emily was how to be able to deliver this product safely without causing that degree of toxicity.

We now have an algorithm that most physicians who are doing CAR cell therapy use to help guide our treatment regimens of when to intervene with tocilizumab to make the therapy safer, when to give steroids if we need to, and when to think about other ways to intervene to make the therapies possible so that they are not in a coma for 14 days and able to have the maximal efficacy without having significant toxicity.  That is an area that continues to be explored as we try to make this therapy safer.

McGinley:       It’s still a treatment that is quite risky, is it not?  Is this not one of the reasons why the FDA and Penn and Novartis and CHOP have all developed a program for it to be used in a limited number; three dozen or so, maybe fewer, medical centers and administered by doctors who are specially trained.  Isn’t that right?  I assume that you have been training some of the doctors?

Shah:               Yes, that is true that we still feel that because of the side effects that many patients like Emily have had with this therapy, we do know a way to safely manage these patients and get them through it, but it requires a lot of specialized care.  And so we all want this to be able to be available to more patients, but we want it to be available safely.  And so it’s really important that we have specialized centers that are adequately trained and really rigorously trained on how to be able to manage and get patients through this as safely as possible.

McGinley:       How are you doing the training?  Are you visiting all of these centers and training in person?

Shah:               There are many sites that have been involved in the trials that led to the FDA approval who have treated patients already and so have received a lot of that training and then there are other sites that are now being added and we do a lot of interacting with those physicians.  Both Dr. Grupp and myself and then as well the team at Novartis who has run these clinical trials, do a lot of one-on-one training with those sites.

McGinley:       Dr. Shah, as you and I have talked about before, one of the challenges of this therapy is that some children do relapse and Emily was fortunate from that standpoint that she was not one of them.  But about 30% of children is that about the proportion of relapse on this therapy and I’m sure you’re doing also work to try to reduce the relapse rate and trying to figure out different ways to use this technology.  What are you doing?

Shah:               So what we hope for CAR cell therapy is that not only will it be effective, but that it will be durable and in terms of durability, Emily leads the field with five years of being in remission.  But not every child who gets CD19 CAR therapy will have that durability of response.  Amongst those patients who relapse after CD19 CAR therapy, some of them are now losing the protein, the actual CD19.  So the CAR cell therapy becomes ineffective.

McGinley:       Is it because the target has—

Shah:               The target is no longer available to be targeted.  And there are multiple mechanisms how that may happen.  What we’ve done at the NCI, we have a long history enough looking at CD22 as an alternate target on the leukemia cell.  So we have a CAR therapy that targets CD22 that has also been found to be very effective.

McGinley:       Yes, and also tell me a little bit about that trial, a little bit more about that trial and also, about the fact that you’re going to move into multiple targets.

Shah:               Yeah, so in terms of the CD22 CAR therapy, it’s very similarly constructed to the CD19 CAR therapy.  You have the CD22 that’s on your leukemia cells and the CD22 CAR that then attacks that [ph].  With that therapy, we’ve seen again, cytokine release syndrome just as you would with the CD19 CAR, but we’ve also seen good remissions and have a patient who is out a few years as well.  The problem again, is the problem of relapse.  We have patients who are similarly relapsing after losing CD22.

So the next trial that we’re going to be focusing on is actually a combination therapy, of a bispecific [ph] CAR that is simultaneous and will target 19 and 22.  And again, the paradigm there is not all that dissimilar from multi-agent chemotherapies.  If you go back to their origins of treatment, of childhood ALL, you started with one therapy and they would maybe have a remission for a few weeks, maybe a few months, if you were lucky.  But then you would have leukemia resistance.  So then you could put into that combination chemotherapy, which is what the current paradigm is of treating childhood ALL.

What we’re hoping to do with the bispecific CAR is have a CAR that is simultaneously targeting both 19 and 22 so that you don’t have this problem of the leukemia and outsmarting the CAR-T cells that is getting us at an individual level.

McGinley:       And what about you?  Are you working on the relapse issue as well?

Maude:            We are.  I think we all feel that while the results that we have seen are remarkable, we always want to do better and we want to be able to overcome the relapses that we have seen and so we similarly are going at the leukemia from a different angle and by targeting CD22 and we also are looking at the other form of relapse, which is having the T-cells not last as long as we would like.  The goal of this therapy is this can really be a living drug and something that can last in the body for months or years and that happens in many patients, but in some patients, that does not happen and the T-cells go away more quickly.

And so we’re studying ways that we can overcome that and make the T-cells slightly differently to see if they can last longer and also add combinations of medicines to try to get the T-cells to last longer.

McGinley:       What about other types of cancers?  Would they be treatable with this kind of treatment?

Shah:               Yeah, I think that’s the goal.  So right now, we started with the b-cell leukemias.  We’re also working on trying to target lymphomas that are also of b-cell origin.  I think for us, that’s definitely an unmet need, thinking about using this CAR cell therapy for lymphoma populations where they may not have other therapies.  And then also shifting to a different type of leukemia altogether so we’re working on developing CAR therapies that target AML, which is acute myelogenous leukemia.

McGinley:       And what about the whole idea of solid cancers?  Is that something that this therapy—and it’s not just Novartis.  We should say there are other companies and other academic researchers all over the country and all over the world who are working very hard on this.  Do you know what the state of play is in terms of—when I talked to Dr. June after the approval, he said that in my studies, the results were that there was a response by a kind of a wide array of cancers, but the big question is how long would it take and will it translate to humans, number one.  And number two, how long would it take to have that happen?

Maude:            Well, in theory, this technology could really target any type of cancer.  If you could find something specific to the cancer that you could train the T-cells to find, this could work across a whole spectrum of cancers.  What has become an issue for solid cancers is really getting access to the tumor itself and overcoming some of the mechanisms that the tumor has developed to try to block the immune system.  And so what we think will need to happen for this to work in other types of tumors is that there would need to be some additional technology to try to overcome some of those barriers.  But certainly, the hope is that it can be expanded much more broadly.

McGinley:       Tom, when you wound up at CHOP and in this trial, were they describing it as a way to get Emily better enough to get a bone marrow transplant?  Or were they saying that it would be hopefully a replacement for the need of a bone marrow transplant?

Tom W:           From our perspective, my wife and I were hoping that it would work and not need a bone marrow transplant but they did say that it would be our call to make after the cells would work depending on how many T-cells were inside her and that as parents, if we wanted to go to bone marrow transplant, we could.  We personally didn’t think Emily was healthy enough to survive a transplant and we had seen many kids suffer through bone marrow transplants and I do tell people now, I was more afraid of significant Emily up for the full-body radiation to go to bone marrow transplant than I ever was of entering here as the first patient in the CAR-T trial.  But we were just hopeful that it would work.

I always felt inside that it would and once it did work—well, we did have that discussion and said as far as us as parents, we did not want to go to bone marrow transplant.

McGinley:       And is the idea that it replaces bone marrow transplant?  How often do children get bone marrow transplants after they get CAR-T?

Shah:               I think it’s too early to say that it will be a replacement for bone marrow transplant because I think right now, we don’t have enough years with CAR cell therapy to know the durability.  And currently, the standard of the curative treatment for patients with multiply relapsed refractory ALL would be a transplant.  But we know that patients are going to relapse after transplant as well.  So I think what the future holds is either some combination where you’re able to use CAR cells as a bridge to transplant, potentially making the transplant less toxic.  If you can get the patient into deep remission or being able to have a way to identify which patient is at risk of relapsing from the CD19 or any CAR-T cell therapy where the transplant would be indicated.

And I think that’s where the future is in terms of where the research really needs to be to make the decision about transplant.

McGinley:       Dr. Maude, we’re going to have Emily join us in a minute but first, I wanted to ask you a question.  When you’re talking to parents about this procedure, this very dramatic treatment, what do you tell them and then what do they tell you after the therapy is over?

Maude:            So when parents are coming to us and they are often really looking for hope and looking for an option after in many cases, having been through many different therapies, we’re very upfront and honest about what to expect and it can be very scary to hear all of the potential side effects.  But we prepare them in a lot of detail and walk them through and I think what they’re relieved to hear or that we have treated many patients and have been able to get them through and when they come out on the other end is coming back months later with their child back in school, playing baseball, doing things with their friends and I have had many parents say to me, “My child hasn’t felt this good in years, since before they were first diagnosed.”

McGinley:       That must be amazing.

Maude:            It is.

McGinley:       At this time, I would love to welcome Emily Whitehead to join us.  Please, let’s give her a round of applause.  [APPLAUSE] Emily, thanks so much.  I hope you’re having a fun time in Washington.  [LAUGHS] You told me you went to the Air and Space Museum this morning.  What was your favorite part of it?

Emily W:         My favorite part I have to say was the planetarium.

McGinley:       What did you like about it?

Emily W:         It was really cool because it looked real.

McGinley:       Now, your father has told me that you don’t remember a lot of the very difficult things that happened to you and I think that sounds like that’s a good thing.  But I’m wondering if you remember how silly he was and the extent he went to try to make you smile and laugh through it.  Like is it true that he—did he really dress up like a Penn State cheerleader?  [LAUGHTER] What did he do?

Emily W:         I remember this one time, he picked me up by my ankles and he held me upside down and helped me do a handstand on my bed.  [LAUGHTER]

Tom W:           My rule was you had to smile at least once a day and as parents, you’re putting this poison into your kids to try to get them better and they’re really just going through hell to try to get better.  So we tried to do whatever we could just to try to get a smile once a day so we would do whatever crazy things we could and sometimes, we would set up the nurses too.  We got a large syringe full of water one time and when Nurse Robb [ph] walked through the door, Emily gave it to them with a water shot.  [LAUGHTER] I got a remote-control helicopter and I’d fly it around the room and Emily would hide under the covers and that was fun up until it got stuck in my wife’s hair.  [LAUGHTER]

McGinley:       She must have loved that.

Tom W:           She made me take away the helicopter after that.

McGinley:       What about the pom-poms?  What about the Penn State pom-poms?

Emily W:         It was really funny when he did it [ph].  [LAUGHTER]

Tom W:           Luckily, there’s no video.

McGinley:       So Emily, you’re back at school now.  You’re in seventh grade?

Emily W:         Yes.

McGinley:       And I’m told that you’re a completely normal preteen child.  [LAUGHS] Tell me a little bit about—now that you feel well, what do you like to do in your free time?

Emily W:         I really like to do any kind of art and I really like to hang out with my dog, Lucy.

McGinley:       Now is it true that you’re a potter?  You took pottery classes this summer?

Emily W:         I did, I took a couple and I’m taking clay class in school.

McGinley:       And you also like to make videos, right?  What kind of videos do you make?

Emily W:         I do.  I make videos for YouTube.

McGinley:       Are they instructional videos or what kind of videos are they?

Emily W:         They’re a bunch of different ones.  There are weird videos that I made with my friends to demonstrate [ph] art things.

McGinley:       Is there where slime is somehow involved in this?

Emily W:         There’s a lot of slime.  [LAUGHTER]

Tom W:           Her channel is Ambassador with an E [ph] because we’re Emily’s ambassadors.

McGinley:       I’ve seen some of the photos of when you returned home to Phillipsburg, Pennsylvania from CHOP and the whole town came out to greet you and it was just amazing.  What was that like to experience that?  You must remember that because that was a good thing, right?

Emily W:         It was amazing because it was raining and everybody still came outside and I remember that my nan [ph] or his mom handed Lucy through the window and she was so excited to see me.  [LAUGHTER]

McGinley:       I bet she was.  Your entire family has really taken it upon yourselves to help other families and other people who are in the same situation and unfortunately, there are probably too many.  Do you hear a lot from different people who are struggling with this?  Different families?

Tom W:           Yeah, so when Emily’s treatment was covered in the worldwide media when it was finally announced in December of 2012, we started getting calls immediately from families from all over the world saying, “We’re being sent home on hospice with our children and we would like help.  And that eventually led us to start our own foundation, the Emily Whitehead Foundation.  But it has really increased ever since the ODAC hearing and the FDA approval.  It’s getting more coverage and now people trust it more now that it has FDA approval and we hear from a minimum of one to two patients per week.  So we take those calls every chance we can get and try to help guide other parents to trust their instincts and make the best call for their child.

McGinley:       And Emily, what happens—I’m told your father was telling me that because you have to go to CHOP so much to be checked out—that other children approach you and other parents approach you while you’re there.  How does that feel?  Do you like that?  Do you not like that?

Emily W:         I do like that because they come up to me and they tell me that I’m an inspiration to them and it makes me feel really good.

McGinley:       So you feel as though you’re really contributing something and actually you’re making a difference.  And what do you tell kids your age or younger who are dealing with this, who just need a little inspiration, a little hope?

Emily W:         I always tell them never give up and always keep believing in yourself.

McGinley:       Tom, you said also that there are very specific things, ideas that you have that you tell parents based on your experience and doctors.  Could you share that with us?

Tom W:           Yeah, and we try to now tell doctors to listen to your patients whenever they’re asking for some of these new treatments.  Not everybody is up to speed yet on these new breakthroughs.  But we also tell the parents that what we noticed when we were there is these doctors have to take care of every patient in the hospital and you only have to take care of your child or your patient.  So we really encourage people to do the research on everything that will be put into your child and to really trust your instincts because if we didn’t trust our instincts and once in a while maybe say no to what was being recommended that day, Emily wouldn’t be with us.

McGinley:       Tell us a little bit about that.  Give us an example of how that worked.

Tom W:           There’s quite a few different examples but for instance, when Emily was in the PICU and she was on the ventilator and they came around to do rounds one morning and my wife, who works in research at Penn State University kept files of everything of Emily and had it on her computer, especially her blood work and some of the residents one morning started talking about Emily’s blood work and her vitals and it sounded like things had taken a turn for the worse and that she needed some immediate care and my wife Carrie was paying attention and she said, “What you just said is not Emily’s blood work.  I paid attention and that can’t be what you just said.  It’s not correct.”  And they actually looked back at the computer and said, “I apologize.  We brought up the wrong patient.”

And we know they’re working around the clock and everyone’s tired sometimes and it just helps to be an advocate and stay on top of things.

McGinley:       Well, and that wasn’t the only time your wife played a very pivotal role in making decisions.  Tell us a little bit about her role.  She’s a researcher at Penn State?  Is that right?

Tom W:           Yeah.  Right now, she’s working in the Military Clearinghouse but she has that kind of training and she’s worked as a project manager in the past on research projects and she was able to always access her work computer and just research all of the different drugs that was being offered to Emily.  And another day, Emily had relapsed and our transplant doctor at Hershey had said, “You don’t have time for a second opinion and we’re going to start the ICE round of chemotherapy, which is three drugs to start with I, C, and E.

And it didn’t feel right to us that day and we just said, “We’re going to go back to the Children’s Hospital in Philadelphia and get another opinion.”  And they said, “She’s not stable enough to transfer for another opinion”, but we turned it down and drove her down there ourselves and had she been given that drug, we were told now that she would have had no T-cells to grow.  So turning it down that day saved her life.

McGinley:       That’s quite amazing.  How many parents would have the confidence in their judgment to do that?  I think that would be a very, very difficult thing to do.  So the message that you give to parents is to truth their own intuition and do their own research as well as listening to the doctors and you tell the doctors to listen to the patients?

Tom W:           Absolutely.

McGinley:       What are the next plans for the foundation?

Tom W:           Our signature event is coming up on October 28th.  We’re having the Emily Whitehead Foundation Believe Ball at the Valley Forge Casino and we’ve invited all of the patients that got the treatment after Emily and we’re trying to get as many together as possible for the first time in one place and bring in a lot of the people.  The doctors are coming and the workers and just have everyone get together and have a night of inspiration and stories and our goal is to continue raising money and help fund growing the T-cells and furthering immunotherapy for the other patients that need it.

McGinley:       Dr. Maude or Dr. Shah, it’s interesting to me and I think to everyone that this was maybe the first cancer drug that was approved first for children.  Is that right?  I think maybe it was.

Maude:            Yes.

McGinley:       So that made it really doubly historic.  Can you explain a little bit about why that is and why children traditionally have been the last to get cancer therapies?

Maude:            Well, it’s often the case that drugs are studied first in adults for a couple of reasons.  One is that we do want to show that the drugs can be safely given in adults before it is moved into children and that’s a big part of it.  But the second reason is that fortunately, pediatric cancers aren’t that common.  And so typically, it doesn’t make a lot of sense for drug companies to really be developing a drug that is specific to a disease with a very small population of patients.  And so this was really amazing that this was developed first in children and first approved in children.

And it’s incredible as a pediatrician to know that the research that we were a part of contributed to this and to know that this is something that is really important and groundbreaking and the studies that were done in children made that possible and I think it’s partially because children are strong, children like Emma [ph] are really strong and being able to study this therapy in children was really what made this therapy advance because we were able to safely administer this in children.

McGinley:       Dr. Shah.

Shah:               Yeah, another part of this is also the medical ethics of children being able to provide consent and so I think that also comes into play when thinking about doing particularly phase one trials that are not intended to look at the efficacy but are really intended to explore the safety and the toxicity profile of these therapies, which is why there’s sort of a scaled down approach.  We start with adults who are freely able to give consent before going to pediatric-aged patients.  But, as the world evolves and we learn more about targeted therapies, I think both with the FDA and I know certainly, even within the pediatric oncology [ph] branch, we are looking at identifying protocols where it does make sense to think about lowering the age limit.  And so instead of having a strict cutoff at age 18, if there is a target therapy and a patient has that target, can you reduce that age limit to start a little bit lower?

When we did our phase one CD22 CAR-T protocol, we were actually able to start at age 16 even though we had to treat adults before children.  And that’s one example of being able to introduce therapies into children a little bit earlier.

McGinley:       Well, unfortunately, we’re out of time but I’d like to thank all of you for joining us today.  A special thanks to Emily for coming down and meeting with us and thank you, Mr. Whitehead, Dr. Shah—

Tom W:           Thanks for having us.

McGinley:       Thank you very much.  We’re now going to exit and I’m going to turn over the

program to my colleague, Fran Kritz for the next panel.  Thank you.

Coding Cancer: When Medicine and Technology Converge

Kritz:               Good afternoon.  My name is Fran Kritz, and I’m a contributing health and medicine reporter for The Washington Post.  This is the concluding panel for today’s program, and we have with us four world-class experts to discuss innovations in prevention, detection, and the treatment of cancer.

Dr. Regina Barzilay is the Delta Electronics Professor at the Massachusetts Institute of Technology.  It’s a computer science and artificial intelligence lab.  Dr. Barzilay’s research focuses on natural language processing and machine learning.  After her diagnosis of cancer three years ago, Dr. Barzilay added a new direction to her research, applying artificial intelligence and machine learning to improve cancer diagnoses and treatment methods in collaboration with researchers at the Massachusetts General Hospital in Boston.

Dr. Otis Brawley is the chief medical officer and executive vice president of the American Cancer Society, where he is responsible for promoting the goals of cancer prevention, early detection, and treatment through research and education.  Dr. Jill Hagenkord, molecular pathologist, is the chief medical officer at Color Genomics, which offers consumers, as ordered by their doctors, genetic testing to help understand risks for common hereditary cancers.

And last but not least, Dr. Douglas Lowy is the acting director at the National Cancer Institute.  Dr. Lowy is a renowned expert in the areas of cancer biology, papillomaviruses preventive vaccines, and tumor suppressor genes.  And thank you all for joining the panel today.  I’d like to remind everyone in the audience and those watching online that you can tweet your questions using the hashtag #PostLive, or leave a comment on the Facebook live stream.

Let’s start with Dr. Barzilay.  Dr. Barzilay, what are you and your colleagues doing that will help physicians use computers to refine cancer diagnoses and treatment?

Barzilay:          So whatever I am doing with computers and machine learning was impacted by my own experience as a breast cancer patient at MGH.  I was treated very well.  But one thing that really strike me is how much uncertainty patients have, and how much all this data that is collected in the hospital is really not used.  And as a patient, I saw that I was an exception.  But once I was treated, I was actually looking at the website of American Society of Clinical Oncologists, and I found statistics that I found to be a travesty.

The statistic says that today, all their [ph] clinical decisions are based mostly on 3% of the populations that participated in clinical trials.  Which means that tons of data that is collected about the patients, people like me who were not part of any clinical trial, is not utilized.  And when the whole process of cancer treatment, and we’ve heard it from the previous speakers, has so much uncertainty.  And it’s really a nondeterministic process where big data can make a huge difference.

So after I came back to MIT after the treatment, I decided I am going to change it.  I was lucky enough to find doctors at MGH, Dr. Kevin Hughes and Dr. Constance Lehman, who also believe that the situation has to be changed.  And together with my student, we started this journey.  So the first things that we’ve done was actually to take big data—and the previous speakers were talking about it—in the area of breast cancer pathologists [ph], which currently is not utilized.  And if you look even at the top journals in pathology, many studies are really, really tiny because now it requires somebody to sit down and encode information by hand to do the study.

So we decided to change it.  We took all the pathologists available from breast cancer from partners, and we translated the free [ph] text into a database.  So now you can automatically [ph] identify all the women who had this condition, maybe in five years recurrent [ph].  You can easily identify the populations that you need and there is no human in the loop.  The database currently has 160,000 reports, which are coded in these 20 fields [ph].  So it’s a very large database.  It spans over three decades, and it’s already used in many studies that look at cancer progression, and the development of a TPS [ph].

Another direction that we are working on, and again, impacted by my experience as a cancer patient, is to better utilize imaging data that we have.  So, currently, you do your mammogram.  It’s a pretty large image.  It’s 3,000 pixels.  It’s read by a human who can only notice limited amount of information that is in this mammogram.  Then it’s summarized into a short text.  And a lot of information is really lost here.  Machines are really good in reading, you know, millions of these mammograms, and answering questions which humans cannot answer today.

Like for instance, to do personalized risk assessment of breast cancer, given specific tissue of the woman.  So you can train machine to take this image and say, “What is the likelihood is five years that the patient may develop cancer?”  Or to see whether the patient, like myself who is on medications, responds to medication or not.  And a variety a of other questions.  Is a woman heading towards recurrence?  So all these data is available.  There are a lot of progress [ph].  I’m sure you all learned about deep learning [ph].  So instead of using deep learning to recognize different types of CADs [ph], we are trying to use it to learn about development of cancer.

Kritz:               Thank you.  Dr. Hagenkord, let’s talk about your vantage point as at the helm of Color.  And that is, what are some common misconceptions that people have about genetic testing, and what can we do to change that?

Hagenkord:     Common misperceptions about genetic testing.  First, it’s important to always be clear what we mean when we say “genetic testing.”  A lot of times, people think I’m talking about the DNA that’s inside somebody’s tumor.  And we do do testing on—Color doesn’t do it—but people do do molecular testing and genetic testing to look at the DNA of tumors.  And that can often tell you how the tumor is going to behave or what drugs it might respond to.  The kind of genetic testing that Color does is looking at all the cells in your body for an inherited predisposition to develop cancer.

One of the common misperceptions about that kind of genetic testing is that it’s deterministic; that if I look at your DNA, I can somehow tell you exactly what’s going to happen, and you’re definitely going to die of this kind of cancer at this decade of your life.  And that’s not really true.  That might seem like a bad thing, but I actually think of it as a good thing, that it’s not deterministic.  If you’ve got a mutation in the BRCA1 [ph] gene, for example, and you’re a woman, we know that you’ve got like a 70 or 80% lifetime risk of getting breast cancer.  So we know that you’re playing a game of life here, and you’ve got loaded dice.  They’re loaded against you.

But because we can actually look at your DNA and see that you’ve got a mutation in the BRCA1 gene, we can actually say, “Hey, dice are loaded.”  And you can ask for some different dice, right?  You can do preventative things to reduce your chance of getting that breast cancer that you’re at 80% lifetime risk for.  So you can do things like just increase surveillance.  Instead of waiting until you’re 40 or 50 mammography, you can start at a younger age, and you can fold in MRI on top of it.  You can take certain medications, or you could even have prophylactic surgeries.  But all of these things are shown to reduce your risk of developing breast cancer and ovarian cancer.

A kind of akin to that concept is that not all genes are created equal, and that we have some genes that when there’s a mutation in those genes, they’re really impactful.  And that would be something like the BRCA1 and 2 genes, or the Lynch syndrome genes, where you know that if you’ve got a mutation in one of those genes, you have a serious risk of actually developing one of those syndromic cancers in your lifetime.  Then there’s moderate-risk genes and low-risk genes.  This is the part of the genome that we’re still really trying to figure out how they work together with genes and environment.

But definitely, the more you get down on this end of the spectrum, where you’re dealing with the moderate-risk genes, something like a CHEK2 or an ATM [ph] in breast cancer, that there are more opportunities to shape your destiny; that genes definitely aren’t your destiny.  It’s neither necessary nor sufficient to get breast cancer, but it does put you at an increased risk.  It empowers you then with the ability to work with your healthcare provider.  And even if, based on your personal and family history, maybe they don’t decide to do any overt clinical action.  There’s a personal utility and a heightened awareness, and being kind of more conscientious about going to your regular population level screening programs.  And then also in your family history, that kind of moderate-risk gene could also drive you to some of those increased surveillance options that I mentioned earlier.

Kritz:               Thank you.  Dr. Brawley, following up on what happens to people who may not even know that they have cancer, may not be being tested, and know about going for testing, in 2012, you published a book titled How We Do Harm: A Doctor Breaks Ranks About Being Sick in America.  And it talks about your frustration with the inconsistency of cancer detection and treatment in the United States.  How do we begin to change that?

Brawley:          Well, number one, we need to realize that everybody deserves good healthcare.  One of the problems that I have, let’s just look at some very simple statistics.  We have had a 35% decrease in the death rate from breast cancer in the United States since 1991.  That’s a wonderful thing.  There are 12 states that have not had a 10% decrease.  Why is it that people who live in Massachusetts have done very well from breast cancer, but people who live in Mississippi and Arkansas, no?  It’s because medicine is not practiced in a regulated way, where everybody gets the care that every human being deserves.

We need to talk about evidence-based medicine, where we practice what the evidence actually says.  This is the most expensive health economy in the world.  We spend $10,000 per man, woman, and child for healthcare in the United States.  The second most expensive country, by the way, is Switzerland.  They spend 6,000 per year.  We spend more on healthcare, and then we have to talk about Massachusetts versus Mississippi.  I can give you the same example and we can use colon cancer instead of breast cancer.  By the way, it’s frequently the same states that haven’t had the 35 to 40% decrease in mortality.

Kritz:               Dr. Lowy, this past Friday, you and your colleague, Dr. John Schiller, received the prestigious Lasker Award for your contribution to the development of the HPV vaccine, which protects against four major types of human papillomavirus, the most common sexually transmitted virus in the United States.  I wanted to ask about vaccination rates of the HPV virus, which have been slower than hoped for in some states.  Why is that and how does that relate to other sorts of treatments for cancer that have not had the kind of uptake you and your colleagues would hope for?

Lowy:              Well, Fran, the first thing that I would say is that HPV vaccination is an area of prevention.  Most of the attention for cancer research really is focused on cancer treatment because people are sick, it’s really important to get them better, et cetera.  But cancer prevention actually has the long-term potential of protecting against many more cases than treatment.  And so it was really extraordinary for me to receive the award, not just for the award, but because of calling attention to the potential importance of prevention of disease.

The uptake of the HPV vaccine has been lower than one might have anticipated, but this is true of many areas of prevention.  Dr. Brawley and I were talking earlier, for example, about tobacco consumption.  We’ve known for more than 50 years that tobacco causes cancer of the lung, many other types of cancer, and cardiovascular pulmonary disease, and yet, many people still smoke.  In terms of HPV vaccination, my expectation is that the new regulations, which go from three doses to two doses, will actually lead to a substantial increase in uptake, so that five years from now, we will actually have a fairly substantially high uptake.

But the problem for HPV disease is primarily in the developing world, where 90% of the cervical cancer deaths occur.  And they are only 3% percent of people in the developing world are actually getting vaccinated.  So our colleagues are starting a clinical trial to see if one dose of the HPV vaccine might be sufficient, because this could be a real game-changer in the developing world because of logistics and cost.

Kritz:               Thank you.  I wanted to ask a question that would relate to all the work that all of you are doing.  Dr. Lowy, you first.  And that is each of you represents different disciplines: machine learning, genetic testing, cancer detection, and treatment.  How does the NCI tie that all together, and how do you all communicate with each other so that as each of you makes progress in your own particular field, it can all be brought together so a patient can benefit; patients across the country can benefit from those disciplines joining together?

Lowy:              We always have room for improvement.  We inadvertently have ended up being siloed because people are focused on their particular areas of research.  And going forward, team science becomes very important.  Interdisciplinary studies become progressively more important.  And we were talking earlier, for example, that using machine learning and deep computer algorithms can be enormously helpful, not just for patients making decisions, but also in terms of increasing our understanding and predictability of who is going to respond to what kind of treatment, et cetera.  Because the amount of data that are accumulating now really is far beyond the capacity of any one person to be able to analyze.

Kritz:               Dr. Hagenkord, how are you connecting with your colleagues, in your field, bringing together research disciplines, that come back to you?

Hagenkord:     In the area of inherited disease testing, there are several ongoing efforts, but there are some that I’m going to call out by name; NIH funded projects that were developed specifically because of this very problem.  We’ve gotten new technology that’s really matured over the last five, six, seven years.  It’s called “next generation sequencing.”  And it’s make the cost of sequencing your DNA just absolutely plummet.

So where it used to be that we would look at your DNA judiciously and infrequently, and usually only when you were very sick, now it’s within rounding error almost free to look at your DNA.  The thing is is that when it gets free then we start looking at not just one gene, or two genes, or 10 genes, or 20 genes.  We start looking at 20,000 genes, plus all of the other rest of the DNA.  And so there’s just a lot of learning that’s going on very quickly.  And as a community, we recognize that we were drinking from the fire hose.  When we were looking at genes just once in a while, in just little spots, we could handle it just fine.  But this new technology really kind of forced us to get organized.

It’s something called “ClinGen” and “ClinVar,” [ph] which, like said, they’re NIH funded.  It’s really the entire community of people, you know, genetic counselors, molecular pathologists, molecular geneticists, clinical geneticists.  And we all kind of go and share our information at this site.  So it used to be when I would be interpreting a genome, and I would see a novel variant—one that hadn’t been reported before, or one that maybe it had been reported before, but I thought that the data was kind of janky—I would call up my colleague in Boston or wherever, and I’d be like, “Hey, so-and-so, I know that you’re an expert in cardiomyopathy.  Have you ever seen this variant before?  What is your evidence for how you classify it.”  That’s how we would work through cases, but that’s just not scalable in the age of genomics.

So now we actually have kind of crowdsourced that onto ClinGen and ClinVar.  And that’s where we collect the evidence for the Disease Gene Association and for the Variant Disease Association.  We’re kind of all agreed upon systematic ways that you can score both of those relationships.  You know, kudos to the medical genomics community for pulling that together.  It’s still in its first few years of existence but shows great promise.

Kritz:               Dr. Barzilay, I’m going to ask you first before I ask Dr. Brawley because I’ll ask the more specific question, are you concerned about having sufficient funding for the work that you are doing?

Barzilay:          Absolutely.  I think that for all people who are sitting here, if you’re thinking where machine learning impacted your life, when you go to Amazon, it recommends you a product.  This is machine learning which makes recommendation.  If you do your history, like your credit history, check again, it’s machine learning.  The only area where there is no machine learning is when you go to doctor.  All these amazing advances in machine learning that happen, they really don’t impact our care, and in particular, our cancer care today.

Today, we can say that from NIH funding, only 3% goes to engineering school, and even a much smaller portion of it goes to computer science and machine learning.  I think we’ve heard not only from me, who is a big advocate of the field, but from many people even before this panel, who were saying how important it is to create personalization.  And this is exactly what machine learning can help you to select the best option to cut the cost of care.

We know that if we severely overtreat, because we don’t know who is going to have really bad complications, who is not, machine learning can predict it [ph].  In fact, for in the case of breast cancer, we just currently finished a paper where we demonstrate for a case of high-risk lesions in breast, where in many hospitals, like MGH, everybody get a surgery, only 13% of women actually had cancer.  So machine learning can help you eliminate all those unnecessary surgeries that are currently used.

And coming back to your point about huge discrepancy in treatment, and I saw this map, and it was painful to see what is the difference between state’s incidents and mortality.  This is really painful.  Here machine learning can really help you if you add automatic reading of mammograms and other things which both decrease the cost of healthcare, and also make the quality more equal across places.  There we really can see a change.

And I hope there will be some change in terms of foundations and government funding, bringing mainstream computer scientists and machine-learning people to change it [ph] the same way as people who changed retail [ph] in ’90s, when there was a Netflix challenge and other things.  And all of a sudden, machine learning became the central tool.  So I wish to see the same thing happening in cancer care and we are very, very far from this date [ph] now.

Kritz:               Dr. Brawley, one of the most challenging jobs in cancer must be allocation.  There are so many different studies.  There are so many different directions.  How do you, at the American Cancer Society—and you’ve had so much experience in government as well—how are those allocation decisions made.  And at the very top level, how do you segment them so that you have enough representation of all the different detection treatment prevention to be able to fund all those streams [ph]?

Brawley:          Well, detection, treatment, prevention, and application of some of the things that we’ve already learned, one of my big messages is we are not applying a lot of the things that we already know.  Some of those things, by the way, can be incredibly cheap and incredibly, incredibly give us huge gain.  We, in our basic science area, fund the lowest-hanging fruit.  And that is the questions that are most easy to ask, and we try to get people to address those questions, and keep a very open mind.  I don’t like to get into disease Olympics, like more money for breast cancer, and more money for prostate cancer.

We have a wonderful drug now in lung cancer that came to us because of lymphoma research in the 1990s that found the ALK gene.  The ALK gene is a kinase.  It’s a lymphoma kinase.  Then thyroid cancer researchers found a drug that blocked the product of the ALK gene.  And then, finally, some lung cancer docs married the two together and we got crizotinib, which the FDA approved about five or six years ago.  Unfortunately, it’s only good for about 2% of people with non-small cell lung cancer, but it’s a really important drug for that 2%.  Here we have thyroid cancer and lymphoma benefiting lunch cancer.  I can show you examples where breast cancer has benefited prostate cancer.  Fund the questions that are easiest to ask, and keep a broad mind about how we can apply them to other things.

Kritz:               Dr. Lowy, that really gives us a great opportunity to talk about the Cancer Moonshot, new initiatives in the field of preventing, detecting, and treating cancer, was passed by Congress in December of 2016, which authorized $1.8 billion in funding.  An initial 300 million has been appropriated in fiscal year 2017.  What examples do you have?  And actually, let me start with, how does that allocation happen?  You have this huge pot of money, and as Dr. Brawley was talking about cancer Olympics, I’m sure that every disease organization could say, “With that amount of money, we would go very far.”  How does the NCI sort through and figure out what are the best allocations?

Lowy:              Great question.  We were extremely fortunate that there was the allocation from Congress.  It’s $300 million a year, actually, over a seven-year period, with the last few years being a little bit less.  But as a result, we have been able to put this year about $140 million worth of research that we otherwise would not have been able to do.  Most of our research needs to be funded over a five-year period.  So you can’t spend all of the 300 million in this particular year in terms of new efforts.

We had a group called our “Blue Ribbon Panel,” which we organized about a year-and-a-half ago.  And actually, Tyler Jacks from your institution, from MIT, was one of the co-chairs because he was the head, at that time, of the National Cancer Advisory Board.  They made a series of 10 recommendations, and these recommendations run the gamut from prevention, all the way to survivorship; prevention, screening, treatment.  I should point out that about seven of those 10 recommendations we’ve started to implement them.

But the reason that we had those recommendations is because of long-term strong support that is bipartisan, from both the Republicans and Democrats, for biomedical research, which got to the point where more investment would lead to more advances.  I would point out that our regular appropriation, that is for the NIH in general, and the NCI specifically, really is far greater than that of the Moonshot itself.  Most of the clinical trials are done with regular appropriation.  All of our basic research, we’ve ramped up the basic research in the last couple of years for NCI, but it still is at abysmally low levels even though we have been fortunate enough to get increases in the regular appropriation.

We fell so far behind because, after the doubling in 2003, for many years, our budget were flat while the investment or purchasing power went down because of inflation.

Kritz:               Thank you.  Dr. Hagenkord, let’s go from the very macro of funding allocation to the very individual of each patient.  The tests that your company produces, they have to be ordered by a physician.  Why is that so important?

Hagenkord:     Good question.  So my company is located in Silicon Valley, and is kind of the heart of this kind of health-tech movement.  Where, you know, if you want to take machine learning and apply it to difficult health problems, that’s what’s going on in the Valley right now.  And Color is one of those companies.  We’ve had the opportunity to kind of watch and learn from other companies around us, and kind of look the successes and failures and struggles.  One thing that we feel like was really important in making a successful health-tech product was not to get the consumer outside the healthcare system, but to integrate the consumer better with the existing healthcare system.

So when we, in our typical Silicon Valley consumer-centric design theory, where you take a problem like buying a pair of shoes, or hailing a taxi, or getting a hotel, and you break it down and you say, “What would be most delightful for the consumer’s experience here?”  Clearly, our healthcare system was not built on this type of consumer-centric design theory.  But if you were to do that to the genetic testing process, what Color has done is looked at what those users are that we want to delight, and it both the physician, or the healthcare provider, and the healthcare consumer.  Because if you don’t close that loop, and you just go and you’re trying to make a useful, satisfying, responsible health-experience for the consumer, but when they take it back to their healthcare provider, the healthcare provider either doesn’t know what to do with it, or doesn’t understand it, or doesn’t feel like they’ve been connected in the process, it’s going to create an unsatisfying experience at the end to that healthcare consumer.

So Color, in every step of their way, has always wanted to reduce the barriers to access to preventative genetic testing.  You know, especially now that the price has gotten so low, to provide this kind of service.  So we want to break down the barriers to access, but are kind of very keenly aware that we’ve got to keep those touch points with the healthcare providers all along the way.

Kritz:               Thank you.  Dr. Barzilay, one of the frustrations I know for cancer researchers is that clinical trial participation tends to be very low.  What I wondered is whether in your fields of machine learning and AI, how will that augment the information that we’re able to get on prevention and on treatment?

Barzilay:          Well, this is an excellent question.  Thank you.  So first of all, let me start by saying that it is true that there are traditional machine-learning methods, do live [ph] on huge amounts of data, which for some cases, may not be available.  But let me just tell you a very short story that I just experienced.

Unfortunately, one of my students got diagnosed with ovarian cancer a few months ago.  She’s 24.  And it was a very rare strain of ovarian cancer.  She got two totally contrasting recommendations from doctors in two leading hospitals in Boston, which was very disconcerting for her.  And then when I asked the doctor why is it so, he told me, “This cancer, only 10,000 women in the U.S. get every year.  We couldn’t even enroll enough people for clinical trial.”  My answer was, “If you take 10,000 women, for 10 years, you have more than enough patients for any complex deep learning methods to learn from it.”  Unfortunately, now this data is totally not used.  So we don’t know.

So my first answer is we need to bring machine learning to utilize the data which is already there, which is just a dead weight in our healthcare system, which we don’t use.  Again, travesty.  The second question, in terms of clinical trial, that’s exactly why NIH and other organization need to bring main machine learning researchers into the field.  Because we do quite a bit of research how to deal with scarce [ph] resources; how to utilize data from multiple sources to help you when you don’t have enough data.

That’s why it’s not enough just to recycle algorithm developed for lipstick recommendation into cancer care, which we currently do.  But to really bring mainstream researchers and change what is done to develop techniques which work for what exactly is needed—to first understand better cancer biology [ph] and to develop more effective techniques for prevention and care.

Kritz:               Thank you.  Dr. Brawley, let’s turn to population health for a moment.  One of your previous NCI hats was as director of the office of special populations research.  Kaiser Permanente has published studies in the last few years showing that when you send patients who might not have been able to take the time off to do a colonoscopy, who might not have had the money for a copay, who might not have had easy access to a doctor, they began sending a stool test which can detect colon cancer in certain circumstances to their homes.  And found that the uptake was fairly high, and that detection went up.  And that, at some point, some rates of the cancer began to go down as well.

How can we connect people who may not have the resources to travel, to have time, who may not know about information?  How can we have examples like this ease of use, and be able to connect to the best of what’s in cancer detection, and get it to more people?

Brawley:          By the way, that’s a grand example of a very inexpensive test, stool testing, which, by the way, the science that supports stool testing in screening for colon cancer is actually a little bit stronger than the science to support $2,000 colonoscopy.  Not to say colonoscopy is bad, but that’s a way of getting things to the people very cheaply, in a large way, and helping a large number of people.

You know, the one thing I would do to help folks, and to try to equalize these problems, and to get this research to people is if I could get every person to have a good honest relationship with a healthcare provider.  It can be a general internist, a pediatrician, a family practice doc, a nurse practitioner, or a physician’s assistant supervised by a physician.  But if people could have a good relationship with someone where they could easily feel free and comfortable to talk about their health problems—those docs, by the way, would need to use computers in order to help them make some decisions—that would go a long way in getting people into the healthcare system, getting them evidence-based medicine instead of some of the practice that goes on now.

I frequently get into trouble for saying this, but we sometimes practice faith-based medicine instead of evidence-based medicine.  Faith-based medicine is the doctor does it because they’re sure it works.  They don’t know anything about the science behind it, but it must work.  We’ve done it for the last 20 years.  Evidence-based medicine is we’ve done the science and actually figured it out.  If we started doing those sorts of things, we’d be oh-so better.  Our system would be cheaper, and we’d have better outcomes, the way the Europeans have better outcomes and a healthcare system that’s cheaper than ours.

Kritz:               Thank you.  To wrap up this really excellent conversation, and thank you each, I wondered if we could do a lightning round, and each of you would tell me, when you come here next year, when we come together again a year from now, what would be something that you would be able to share with the audience that was new in your field?  Something that is just beginning now, and that might be just be either the glimmerings or advanced results in whatever discipline; whether it’s more communication; whether it’s ground-breaking research; whether it’s a hope for some things that are starting to percolate.  Each of you would talk about the work that you’re doing that is exciting.  Dr. Lowy?

Lowy:              Well, I would pick cancer screening.  I think we are going to end up in an era where the precision of cancer screening will go up.  The range of cancers for which we will have cost-effective testing will also go up.  It won’t be for everything, but we will see incrementally improvements in this area, and some of them may be dramatic.

Hagenkord:     So as luck would have it, today Color launched a $149 BRCA1 BRCA2 test.  And a few years ago, that same test would have cost almost $4,000.  We’re actually running a promotion for breast cancer awareness month to do it for $99, with $50 testing of at-risk family members.  It’s a physician-ordered test.  It’s a clinical-grade test.  You get access to board-certified genetic counselors every step of the way to help you understand the results, and help guide you to your next steps, and to work with your physician to make a preventative plan if you get a positive.

And so what I would like to see is—and this is a quote actually from Dr. Mary-Claire King, who recently wrote an opinion piece where every woman over the age of 30 should be getting BRCA1, 2 testing, that in 2017, and definitely in 2018, there’s no reason that any woman with a BRCA mutation should be dying of breast and ovarian cancer.

Kritz:               Dr. Brawley?

Brawley:          To take off on Dr. Lowy, I think we’re going to have more screening tests.  We’re going to have a blood test looking for circulating DNA fragments that might indicate a person has an early cancer.  I just hope that we actually validate these tests to actually show that they save lives before we start implementing them in the population the way we have with screening tests in the past, and then 20 years later realize that screening test actually doesn’t save lives.

Barzilay:          What we are currently doing, we’re already demonstrating that using deep learning and the raw data that we’re collecting, like mammograms or pathology [ph] slides, can already make some really important predictions.  What I would like to see in a year, and I believe we can do it in a year, is to provide full probabilistic [ph] progression of the disease so that we can reliably say for women, who is going to get cancer, need to be screened differently.  Who doesn’t need to be screened because they are likelihood is so low there’s no point for them to come every year.  And for people who unfortunately got cancer, to provide, again, a likelihood of recurrences, and the sensitivity to treatment.

Kritz:               Unfortunately, that’s all the time we have for today.  I’d like to thank our panelists, Dr. Barzilay, Brawley, Hagenkord, and Lowy, for participating in this really wonderful discussion, and to everyone in the room, and those watching online.  To find clips from today and on more upcoming programs, please visit  Thank you all.  Thank you.

M/F:                Thank you.

Sponsored Segment: American Association for Cancer Research CEO Margaret Foti Presents AACR’s Cancer Progress Report 2017

Foti:                 Good afternoon, everyone.  It’s really wonderful to be here.  It’s very exciting to be here at this Washington Post event, and a great honor to share the stage with distinguished speakers, many of whom are members of the AACR.  And I wanted to especially mention two leaders from the government sector: Dr. Douglas Lowy, acting director of the NCI, and Dr. Scott Gottlieb, who’s the new commissioner of the U.S. FDA.

The AACR really appreciates the opportunity to help increase public understanding of the myriad diseases we call “cancer,” and to briefly outline the dramatic strides we are making against cancer.  Formed 110 years ago, the AACR is the world’s first and now the largest professional organization in the world that is dedicated to conquering cancer by accelerating progress in every area of cancer research, from cancer biology to early detection, diagnosis, treatment, and prevention.

Research is absolutely pivotal to the vital mission of preventing and curing all types of cancer.  It provides us with a deep molecular understanding of how normal cells work, and how and why they become cancerous.  As you will hear today from others on the program, this growing knowledge, which is powered, in large part, by the exciting new technologies available today, is allowing us to devise better approaches to cancer prevention, increase the number of cures through new developments and treatment, and markedly reduce cancer rates.

The important research that accelerates progress against cancer is made possible only because of the generous federal funding for the National Institutes of Health and the National Cancer Institute, which, after expert scientific review is then distributed to meritorious research projects conducted at comprehensive cancer centers and institutions all over this country.

Also, in the regulatory arena, sufficient funding for the extraordinary work of the FDA, undergirded by the latest cancer science and medicine, is absolutely essential for the rapid delivery of state-of-the-art discoveries to patients.

So now in its seventh edition, the widely-read AACR Cancer Progress Report really provides a snapshot of progress against cancer, and I’m really thrilled to share with you today some of the advances detailed in the 2017 report, which was really unveiled just last week in a congressional briefing here in Washington, D.C.

The enormous progress described therein, along with the deeply personal stories of patients, who are not only alive today but who are also enjoying quality of life, really underscores why our nation must continue to make federal investments in cancer research and biomedical science a top national priority.

In the United States, the overall cancer death rate has declined by 25% since 1991.  And more and more people are surviving longer a cancer diagnosis.  And the number of cancer survivors in the U.S. has grown to 15.5 million.  Among the most exciting recent advances is the emergence of two new types of cancer treatment: targeted therapy and immunotherapy.  Targeted therapy refers to drugs that target molecules that have been discovered to play a role in causing cancer and driving its progression.  These drugs target cancer more precisely than cytotoxic chemotherapies, and therefore tend to be more effective and less toxic.

In the 12 months covered by the 2017 report, seven new targeted therapies were approved by the FDA.  As illustrated in the report, these drugs are bringing new hope to patients like Terry Woodhall [ph] and Evan Freiburg [ph], who turned to this type of novel treatment after diagnoses of ovarian cancer in Terry’s case, and advanced leiomyosarcoma in Evan’s case.

Immunotherapy is now having a dramatic effect in patients with certain types of cancer.  Immunotherapy refers to treatments that work by unleashing the power of a patient’s immune system to fight his or her cancer, much like the way the immune system fights pathogens or viruses that cause the flu.  Our report focuses on a group of immunotherapies called checkpoint inhibitors, which work by releasing the brakes on the innate cancer-killing power of our own immune system.  This type of treatment is improving survival and quality of life for patients who have been diagnosed with an increasing array of cancer types.

Adrian Skinner [ph], Bill McCone [ph], and Carrie Best [ph] are three of the many patients who are benefiting from treatment with a checkpoint inhibitor.  Adrian, Bill, and Carrie, who were diagnosed with metastatic cancer in 2013 and 2014, are currently living full lives with no evidence of disease, long after they were told they had just a few months to live.

And, as you will hear later, just a few weeks ago the FDA approved the first of a revolutionary type of immunotherapy called CAR T-cell therapy.  Clearly many more transformational breakthroughs in cancer treatment and prevention are needed.  The public health challenge posed by cancer is unfortunately growing.  In the United States alone, it’s estimated that the new number of new cases will rise to 2.3 million in the year 2030.  This rise is projected because cancer is primarily a disease of aging, and the segment of the U.S. population at age 65 or older is, as you know, growing every year.

Also we need to make sure that everyone will benefit from groundbreaking advances being made against cancer.  Cancer can strike any one of us—no age, gender, race, ethnicity, or socioeconomic status makes a person immune to this devastating disease.  Sadly, the burden of cancer is shouldered disproportionately by certain segments of the population, including racial/ethnic minorities, patients of lower socioeconomic status, and other groups in this category.  This is completely unacceptable and it’s therefore imperative that greater efforts are made to reduce and eliminate cancer heath disparities.

To address the daunting challenge of a cancer-free world, we must work together more to speed the pace of progress in cancer treatment.  In the report, we discuss how increasing our understanding of cancer by sharing data and harnessing the power of big genomics and clinical outcomes datasets has enormous potential for future progress.  An important example is AACR Project GENIE, which brings together and provides public access to such data from institutions in the U.S. and abroad.  Increasing collaborations among stakeholders in the cancer community are really vital if we are to make further rapid strides in our quest to prevent, intercept, and cure all cancers.

As noted in this last slide, the AACR already had long-standing relationships with sectors in the cancer field.  As you can see, academia, government, industry, societies, philanthropic and advocacy groups, and, increasingly, payers and the lay public that have led to meaningful, life-changing advances for the benefit of patients.

Earlier this year, despite the significant budgetary and political challenges of our nation’s policymakers, we were provided at the NIH a wonderful increase of two billion dollars in annual funding increase for FY 2017.  Now another two-billion increase for the NIH is on the table for consideration in FY 2018, and we urge Congress to make this proposed funding a reality before the end of this year.

With a strong bipartisan commitment from Congress to make investments in biomedical research a national priority, we will be able to achieve more progress, inspire more hope, and save more lives from cancer and other human diseases, not only today but for future generations.

So thank you very much for the opportunity to offer the perspective of the AACR on chasing cancer, and for your kind attention.  I hope you enjoy this wonderful program ahead of me.  Thank you very much.

Sponsored Segment: Medical Experts Discuss: Are We Finally at the Tipping Point in Cancer?

Nelson:            …Thrilled, if not excited, about what we just heard in this progress report and what the future looks like.  And I’m struck, as we’re building these new targeted therapies, now moving into immunotherapy, just what an impact some technologies have had in how we’ve moved forward, particularly genome sequencing technologies.  I mean, we’ve known forever that cancers were diseases of acquired defects in genes; now we can inventory all these defects and use them to help us build new treatments, to get the right treatment to the right person.  Where’s technology going to take us next?

Chen:               Well, you know, it is a really amazing time for the field of cancer research.  And I think largely it’s built upon over three decades of molecular biology, right?  The seeds that were set back decades ago have now led to a point where we understand the cancer process better, why people get cancer, how these cells function, but we also have more therapeutic tools than we’ve ever had in the history of this field.

And so I think we’re at a point today where we’re seeing that emergence of understanding of science and the development of new drugs to help patients with cancer.  And so a really nice example of that, I think, was just highlighted around the emergence of cancer immunotherapy and targeted therapies.  And when we think about these two technologically-different approaches, they’re really interesting.  So a lot of the targeted therapies, they focus on understanding of a particular pathway: a patient that has cancer that walks into the doctor’s office and has a cancer that has a very specific defect.  And you can target a drug to that defect.  And when that’s done well, you see spectacular results.  But the downside to targeted therapies can be, A, it doesn’t for everybody—it only works if you have a very specific mutation generally—but also that the therapy is static, meaning if that cancer cell somehow evolves away from that pathway, that drug’s no longer going to work.

And that’s where I think that we’re starting to see in the field this marriage between very specific targeted therapies, with a different approach, like cancer immunotherapy.  In cancer immunotherapy, we generally are trying to stimulate a patient’s body to recognize and attack the cancer.  And, unlike a static approach, your immune system’s built to interact and find things that are foreign, and so if your cancer looks foreign to your body, your immune system can start to attack it from all sorts of different angles, which makes it more difficult for those cancers to escape the therapy.

So it’s an exciting time, and I think one of the things as a field we need to figure out is, we have all these disparate approaches to the treatment of cancer; is there some way that we can start to put these things together that makes sense scientifically and will really increase what’s possible for patients who are fighting this?

Nelson:            Well, one thing that trikes me with that is that we’re beginning to be able to measure many of these things in many different dimensions.  As you know, years ago, when someone would come to us with a cancer for treatment, we knew their name, their age, generally the diseases that ran in their family, their general health state, and a pathologist had looked through the microscope and told us what kind of cancer they have.  Now we have gene sequences, we have huge troves of data in electronic medical records and the like.  How are we going to bring these large amounts of disparate data together?  Where do you see big data science going in this problem?

Chen:               Well, like you, I love the idea that a patient that has cancer, a diagnosis of cancer today, will be able to walk into a doctor’s office of the future and really know, at a very deep level, what’s driving that cancer.  Why did that cell become a cancer cell?  And why isn’t the immune system already destroying that cancer?  And so that idea that you could do that, just like today you know what year a patient was born and what their family history is, I believe that very soon—and it’s already emerging now—that we will have that kind of information so that you know what’s driving that cancer and then you can pick the right therapy or, in some cases, it’s going to be a cocktail of therapies meant just for that patient.

Now, I think that it’s an interesting time not only with the emergence of these different therapies, but technology is changing.  And we’re at this interesting intersection between our understanding of cancer biology and the development of really novel technologic approaches.  And the kinds of things that I’m talking about are things like machine learning.  All of us, as scientists, when we were training, we learned to ask very specific questions—essentially what we call a reductionist approach.  We try to take complicated problems and simplify them down into something we can understand.

Nelson:            Like putting a cancer cell on a dish rather than someone’s body.

Chen:               Exactly.  And trying to isolate the one simple question we want to answer.  And that’s partially because that’s how our minds work.  We can answer very specific questions when you narrow it down.  The beauty with machine learning is machines think in a completely different way, and machines can find associations that for us, as humans, we just can’t see.  And so when you put those two things together, you have the possibility of really making leaps that we just don’t understand yet.

Nelson:            Do you think this is going to be the way that the next generation of immunotherapies are going to become understood or developed?  I mean, the environment around the cancer—as you mentioned, with targeted therapies, we were kind of only thinking about the cancer cell like it was all alone, but there’s many different immune cells, many different infrastructure cells, and to get a handle of what all of them are doing at once in an ecosystem, if you will—is this the kind of thing you’re talking about?

Chen:               Oh, I think that’s exactly it.  The immune system is really complicated.  Even when you look at the immune response just within the confines of a tumor, there are so many different cell types, and these cell types have different states that I think for us, as scientists, as humans, it’s hard to understand all of the different interconnections that are happening simultaneously.

But the kinds of things that we can start to think about in terms of machine learning relate to, “Okay, well, how many different cell types are there?  How many states are there?”  And, to take it one additional layer, “What is the relationship between them?  Are two cells interacting close together?  Are two cells sending signal across a bigger distance?  And what does that mean?”

Well, I can tell you I have trouble even conceptualizing it, but when you can feed that kind of data into a machine, the machine can start to figure out those complicated associations.

Nelson:            So you’ve been around in academia and biopharma and the like all throughout our industry.  To maximize how well we can translate these new opportunities into new approaches to diagnosis, detection, prevention, and treatment, we’re going to have to think about connections between institutions, between agencies.  How is this going to work?

Chen:               Well, that’s a big problem.  When you take an example like machine learning, machine learning generally requires big data sets, big data sets that are organized in a way that the machine can actually read across them.  And that’s not actually a trivial challenge to solve.  It takes a lot of work upfront to set it up.  And ultimately this will require a level of collaboration across a number of different groups, all the different groups working on cancer research today.

Nelson:            The government, regulatory agencies, funders, all of them.

Chen:               Exactly.  And if you think about organizations like AACR, if you think about organizations like the National Cancer Institute, or if you think about organizations like the FDA, these are bodies that have the ability to start to influence that collaboration, and we really need it.  A nice example: just last week, the FDA hosted a very important conference that brought together all of these organizations to try to answer the question of, “How are we going to move ahead with diagnostics in cancer immunotherapy?”  An incredibly complicated question, but if we can really start to get that commitment, that commitment across companies, across government agencies, and across academic institutions, I think that’s a really, really important step.

Nelson:            Well, I, for one, think we’re going to get that commitment, and that the 8th annual progress report is going to be even better than this one.  Thanks, Dan.

Chen:               Thank you.

Nelson:            All right.  Turn it back over to the Washington Post.