The Peter Attia Drive - #32 - Siddhartha Mukherjee, M.D., Ph.D.： new frontiers in cancer therapy, medicine, and the writing process

00:00:00.000 Hey everyone, welcome to the Peter Atiyah Drive. I'm your host, Peter Atiyah.

00:00:10.140 The Drive is a result of my hunger for optimizing performance, health, longevity, critical thinking,

00:00:15.600 along with a few other obsessions along the way. I've spent the last several years working with

00:00:19.840 some of the most successful, top-performing individuals in the world, and this podcast

00:00:23.600 is my attempt to synthesize what I've learned along the way to help you live a higher quality,

00:00:28.360 more fulfilling life. If you enjoy this podcast, you can find more information on today's episode

00:00:33.000 and other topics at peteratiyahmd.com.

00:00:41.280 Hey everybody, welcome to this week's episode of The Drive. My guest this week is Sid Mukherjee,

00:00:48.300 who is a remarkable writer, in fact a Pulitzer Prize-winning writer, and a remarkable physician

00:00:54.820 and scientist. In fact, probably there's nobody I know that combines those three things

00:00:59.520 as efficaciously as Sid does. His biography reads like I'm making it up. He studied biology at

00:01:06.560 Stanford. He then became a Rhodes Scholar, went to Oxford, earned his PhD in immunology, returned to

00:01:12.140 the United States to earn his MD at Harvard, etc., etc., etc. Fast forward to today, he is an associate

00:01:18.620 professor of medicine in the Division of Hematology and Oncology at Columbia University,

00:01:22.960 which is where we met to do this interview. He has published also and continues to publish

00:01:27.760 consistently in both the New Yorker and the New York Times, which in and of itself is quite a

00:01:32.300 distinction as I would learn. Typically one is on either side of those, but not both. And of course

00:01:37.600 he's published in the New England Journal of Medicine, Nature, in addition to a whole host of

00:01:41.380 other medical journals. I met Sid probably five years ago at a dinner that was set up by Lou Cantley,

00:01:47.500 someone I'll be interviewing very shortly, who will be a guest obviously soon. And while I remember

00:01:52.400 that dinner very well, I was surprised to learn that it left such an impression on Sid. And he

00:01:57.040 described it as something to the effect of one of the most interesting and perhaps important

00:02:00.340 scientific collaborations in his life that stemmed from it. As we kind of jotted out a napkin experiment

00:02:06.340 that went on to become a paper that was published by a group led by Lou and Sid. And that paper was

00:02:13.940 published this past summer. And we talk about that in detail. That paper involved the use of ketogenic

00:02:18.700 diets in combination with a class of drugs called PI3 kinase inhibitors. We're going to go into great

00:02:25.200 detail on that. So I obviously don't want to repeat any of that stuff here. But I think for those of you

00:02:29.160 that are interested in cancer, you're obviously going to find this episode very interesting. But

00:02:33.440 the other thing with Sid is it doesn't matter if you have no interest in cancer. I think you'll find

00:02:38.060 this discussion interesting because Sid has a way of making everything interesting. And that to me

00:02:42.320 is part of Sid's gift. When I got his book, The Gene in the summer of 2016, when it came out,

00:02:48.720 it's one of probably only six books in my life that I was not able to stop reading from the moment I

00:02:56.260 started. So it was one of those things where everything else I was doing had to be put aside

00:03:01.400 for a few days until I could finish that book. That's just the way Sid writes. And that's also the

00:03:07.480 way he speaks. He is a unique human being. And I think that will come across in this interview.

00:03:13.300 The other thing that was a total pleasant surprise to me was in doing the research for this podcast

00:03:17.740 was coming across a book that I was ashamed to admit I didn't even know he'd written called

00:03:22.180 The Laws of Medicine, or I believe it's called The Three Laws of Medicine. And we talk about these

00:03:28.080 three laws. And rather than even stating them now, I just think it's worth, this podcast is worth the

00:03:33.020 price of admission just on the basis of understanding those three laws. So with that, I hope you will

00:03:38.480 welcome Sid to the show. And I do want to just remind folks to please sign up for the email list.

00:03:44.860 I've been putting a lot of effort into those emails every Sunday morning, they go out. And I hope that

00:03:49.920 they're at least worth some value. I think they are. And I like to be able to kind of share things

00:03:53.700 with folks that I'm reading or seeing along the way. And they don't always have to do with longevity.

00:03:57.380 Keep in mind, one of the things that I think takes up more actual time than anything else with respect

00:04:03.680 to this podcast is putting together the show notes. So Bob and Travis work really hard on those. And

00:04:09.260 the feedback we've been getting is incredible. People keep saying, my God, how do you make these

00:04:13.420 things? And the short answer is, I don't. I don't do any of it. But Bob and Travis do, especially

00:04:17.340 Travis. And I think that if you spend a few minutes looking at that stuff, especially if you find some

00:04:22.080 of the content challenging. And when we get into technical terms, which we do on some of the podcasts,

00:04:26.240 you're pretty much going to find everything in the show notes. Lastly, if you're enjoying this,

00:04:30.000 it would be an honor if you would head on over to Apple Podcast Reviews and leave us a review,

00:04:34.580 especially if it's a positive one. But we'll take a negative one too, as long as you can be

00:04:37.880 constructive in your feedback. So without further delay, here is my guest today, Sid Mukherjee.

00:04:46.700 Hey, Sid. Thanks for making time.

00:04:49.360 My pleasure. Thanks for being here.

00:04:50.800 Yeah, I don't get up to this part of the city very often. It's a bit of a hike.

00:04:55.000 Well, it's a massive medical school and it's hard to imagine it anywhere else except for

00:04:59.520 uptown in this way. We go right all the way to the river.

00:05:03.260 It's amazing. And the last time I was up here was to see another one of your colleagues on the

00:05:06.760 other side of the street, Rudy Leibel, who's a good friend. And I used to be up here a lot more

00:05:10.580 often. So it was nice to come back. Most of our podcasts go really, really long. This one,

00:05:15.060 I don't think we have that luxury of time. So I kind of want to get right into things. But

00:05:18.880 before we do, I certainly think for the listener who doesn't know you well, your background,

00:05:22.880 which I'll allude to a lot in the introduction. So we don't spend too much time on it. But you

00:05:26.680 grew up in India, came to the US. Did you do college here?

00:05:30.560 I went to college at Stanford. Yeah.

00:05:32.360 Oh, that's right. You went to undergrad at Stanford. Okay. Then medical school at Harvard.

00:05:35.500 Actually, then in the middle, I was away for three years. I did...

00:05:39.280 You did a Rhodes Scholarship. Yeah.

00:05:40.400 Or you're a Rhodes Scholar. Yeah.

00:05:41.320 I was a Rhodes Scholar. And that's where I got my PhD. And my PhD is in immunology,

00:05:44.940 which is a subject that I left behind, went to medical school at Harvard Medical School,

00:05:51.080 then did my fellowship, my internship, my residency at Mass General Hospital,

00:05:57.260 fellowship at the Dana-Farber Cancer Institute, and then started my own lab and clinical practice

00:06:02.880 at Columbia University. And have come back to immunology in a strange, widening circle of a way.

00:06:10.060 Yeah. The first time we met was a dinner that Lou Cantley had planned for us. This is about three

00:06:15.780 or four years ago. And I remember at the time, the topic that Lou is passionate about that I'm

00:06:20.960 passionate about was sort of metabolism of cancer. And at the time, it wasn't something that seemed

00:06:25.120 as interesting to you as it is today. And I know today I want to talk so much more about that because

00:06:28.860 the work you guys have done in the last few years is in many ways what I think is the most

00:06:32.420 interesting stuff to talk about. It was also a tough dinner because you don't eat most things.

00:06:36.580 It's hard to have dinner with someone when half the menu is off the menu. So anyway,

00:06:42.660 we somehow managed and scraped by and it was a wonderful evening. It actually led to,

00:06:47.080 what was that, five years ago?

00:06:48.520 Yeah.

00:06:49.160 Led to one of the most interesting and perhaps important scientific collaborations in my life.

00:06:54.260 With Lou.

00:06:54.780 With Lou. But it was that dinner which we kind of hammered. It was in a Japanese restaurant on the

00:06:59.700 Upper East Side in a tiny little place.

00:07:01.400 You remember? That's exactly, I know exactly where we were.

00:07:03.300 And exactly where it was. Well, because we wrote on a, you know, it was a little bit like one of

00:07:07.340 these napkin experiments where you write on a napkin an idea and that idea takes five years to come

00:07:13.360 alive. This thing that was sketched on a napkin that evening and is now leading to a actually kind

00:07:18.920 of a massive clinical trial across multiple sites. Very energetic teams coming into all of this out of

00:07:25.840 that little napkin on a Japanese restaurant.

00:07:27.940 Yeah. That was a really fun night. Prior to that, I had read The Emperor of All Maladies. I don't

00:07:32.480 think The Gene was out yet.

00:07:34.080 The Gene was not out yet.

00:07:35.000 I want to spend just a couple minutes on The Emperor of All Maladies because if there's anybody 0.98

00:07:38.000 listening to this who hasn't read it, you won the Pulitzer Prize for that book, I believe.

00:07:41.220 I did, yes.

00:07:41.760 Yeah. So it's a must read. Having myself studied in oncology, there was so much that I learned.

00:07:48.460 You know, I trained at Hopkins, which is, you know, so we're in the Halstead school of where

00:07:52.900 the mastectomy was created, where many of these things were created. But to really understand the

00:07:56.980 history of Bernard Fisher's role in the mastectomy, it's just an unbelievable story. My only criticism

00:08:02.820 was at the time quite a depressing story. I mean, I'm not saying that to be critical. It's less a

00:08:09.100 criticism of youth in the field.

00:08:10.460 No, no. It was, to me, actually, interestingly, people often bring this idea up. For me, it was

00:08:15.060 actually, it's far from a depressing story. It's just the opposite. In fact, if we don't contend with

00:08:21.180 the question of how, I mean, The Emperor of All Maladies, just for people who don't know, is a

00:08:25.840 history of cancer, starting from its first description in Egypt, right down to my own patients,

00:08:32.400 thousands of years of a journey against a disease that seems to morph and change over time.

00:08:38.060 And every time we look at it, it has a new form. It reflects our own diversity, to some extent,

00:08:43.340 our own wiliness, our own imagination as humans. So to me, not a depressing book, because it is a

00:08:49.860 way to look directly at the face of the enemy. And I don't find that depressing. I find that

00:08:54.680 clarifying. And there are many, many high points in this journey. There's the invention of the great

00:08:59.880 surgeries that save tens of thousands, hundreds of thousands of lives around the world, the dramatic

00:09:05.620 advances against breast cancer, and most importantly, against some variants of leukemia, where the

00:09:11.600 mortality was 100% in 1950, and is 5 or 10% now. 95% change in mortality is a huge difference in

00:09:21.120 must rank as one of the great medical inventions of our time. So for me, far from a depressing book,

00:09:26.400 but to me, a clarifying book that tries to clarify why we're here today, where we're going, what happens

00:09:32.160 next, why we're not doing certain things, why we are doing certain things. So that's my impression

00:09:37.200 of what happened at the end of that book. When did the idea to write, as it's really referred to,

00:09:42.080 the biography of cancer, when did that idea come to you? Was it in your fellowship, in your training

00:09:46.360 at some point? Yeah, I was training, and it was a very simple moment. Actually, I remember a woman

00:09:50.900 who I was treating for cancer came to me, and I was giving her yet another trial of targeted therapy,

00:09:57.820 a new kind of therapy. And she finally sat down one afternoon, and she said, where are we going?

00:10:02.320 With all of this, why are we here? How did we get here? And she was, of course, asking it in a very

00:10:06.980 personal level, but you could take that question and make it a much, much larger question. Where are

00:10:12.020 we going in this battle against cancer? Why are we here today? What happens next? Why aren't we

00:10:17.800 elsewhere? How much of this is the wildness of this family of diseases? How much of this is the

00:10:22.960 capacity to use the greatest skills of our imagination against this illness? You know, in the 1930s,

00:10:30.760 1940s, 1950s, when you asked a child what the outer limit of their scientific imagination was,

00:10:37.160 they would say, I want to be a rocket scientist, and I want to send a rocket to the moon. By the

00:10:41.700 1950s, 60s, 70s, if you asked that same child what the outer limits of their scientific imagination

00:10:48.160 would bring to the world, they would say, I want to cure cancer. It began to define the limits of our

00:10:54.320 scientific prowess, or the limits of our imagination, a world without cancer. So there is a sense in which

00:11:02.000 this is such an elemental illness. So much of our culture is now defined through the lens of cancer.

00:11:09.160 And what was shocking as a fellow, when I started encountering cancer in a clinical sense,

00:11:14.540 what was shocking was that there was no such history, that it was all ad hoc, and we knew

00:11:19.520 little bits and pieces of it was like looking at the enemy through a patchwork quilt with little

00:11:24.320 holes in it. And the attempt here was to say, well, what is the full story? What does the story look like?

00:11:29.720 When did this start? Why did we end up here? What happens tomorrow? What happens way into the

00:11:34.740 future, a hundred years? What will cancer look like? And a lot of thought experiments go into

00:11:39.540 the book. So that's sort of the genesis of the book, and that's how it came about.

00:11:44.280 The research is also remarkable for a book like that. Anyone who's read it will appreciate. And again,

00:11:49.260 I think what's nice is this is one of those books where you can be an oncologist and read it and find

00:11:53.800 it staggering, and you can be someone who has lost a loved one to cancer but wouldn't know the

00:11:58.900 difference between a sarcoma and a leiomyosarcoma. It doesn't matter. The book resonates, which

00:12:03.980 obviously speaks to your ability to tell a story. And that's sort of, to me, what's mind-boggling

00:12:08.800 about that book is the way, and it's the same in the gene, by the way, is that you weave in and out

00:12:12.740 of a personal story and then something that's very dense scientifically. And on a personal level,

00:12:17.180 this is challenging because I'm in the process, as you know, of writing a book and doing a pretty

00:12:20.360 lousy job of it, I think. But it's this challenge of you want to be able to do the science

00:12:25.640 justice, but you need to be able to tell a story. So how does the scientist, Sid Mukherjee,

00:12:32.320 get along with the writer, Sid Mukherjee?

00:12:35.540 Well, I don't think they're two separate people. I think they're integrated into one person,

00:12:39.680 and that's very important. There are many people in my world, they wrote as a process of thinking.

00:12:47.360 Stephen Jay Gould comes to mind, not to draw a ludicrous comparison. Charles Darwin wrote to think.

00:12:52.100 Oliver Sacks wrote to think. I would suspect that Atul Gawande writes to think. I mean,

00:12:56.460 I know Atul. So the two people are fundamentally not different people in my brain. In order to do

00:13:04.020 the scientific work that I do, I need to think it through, often through the essays that I work on.

00:13:10.120 And they inspire in a kind of yin-yang or roundabout way, circular way, ways to find new ways of thinking

00:13:17.060 about the world, the cancer world. So there isn't a conflict. I don't feel a conflict.

00:13:21.140 The process is probably of interest. I mean, the process is that when I started writing Emperor,

00:13:26.700 I sort of made a personal vow or a strategy in writing that this was my first book. I'd never

00:13:32.620 written a book before. I wrote down some principles or tenets that I would follow,

00:13:37.280 and I've kept to them in every book since. The first one was that there will be no scientific

00:13:43.520 abstraction. At no place or point shall you go through five pages or three pages without there

00:13:50.140 being a human being in the middle of this. I'm a translational researcher. I'm a human scientist.

00:13:55.700 And that meant that I've made a personal promise that you wouldn't go through five pages without

00:14:00.600 understanding what the payoff of these pages that you've really worked your way through,

00:14:06.520 often as a reader, what the payoff is. So if you look carefully through that book,

00:14:11.240 every five to seven pages, the story comes alive in a human story,

00:14:15.200 in a scientific story, and in a scientist's story. They all intersect. And sometimes through

00:14:20.720 my stories, there are times that are tough. And I'm writing about the description, the first

00:14:25.580 cancer-causing genes, oncogenes, and their mutations. It's tough to say, well, what's the

00:14:31.300 human being? This is a laboratory experiment in which you sprinkle tens of thousands of bits of

00:14:37.160 genetic material onto cells and ask the question, which cells turn from normal to cancerous? And

00:14:42.780 that's the way you trap one of these cancer-causing oncogenes. What's the human story behind it? Well,

00:14:48.440 there are two stories. One story is quite lovely. It's the story of Bob Weinberg, the scientist who

00:14:53.820 you know, walking through Boston in a snowstorm and suddenly realizing, not in a one-to-one manner,

00:15:02.040 but this idea of sprinkling tens of thousands of genes onto pieces of genetic material onto cells,

00:15:08.480 like a snowstorm of genes, there's a kind of congruence to that story. It's not like Bob

00:15:12.700 Weinberg woke up one morning and saw a snowstorm and said, that's the experiment I should do. But

00:15:16.760 there's a kind of emotional congruence to the backstory of a scientist. But then the second story

00:15:24.440 that illustrates this point is to walk through a patient telescoping down, or rather microscoping

00:15:32.080 downwards from their outer cancer, a tumor, a lump, a mass that is about to kill them, a real patient

00:15:38.860 of mine, a man with lung cancer. And slowly in that same story of this man's illness, begin to

00:15:47.340 microscope down to the fact that he actually has in his cells this mutant gene that Bob Weinberg once

00:15:53.720 caught in this snowstorm of sprinkling genes on cells. And all of a sudden, this man's cancer

00:16:00.120 is sitting in a room in Boston, surrounded by his family, dying of metastatic lung cancer.

00:16:06.440 But at the heart, at the root of that lung cancer is that very same gene, that very same oncogene

00:16:12.220 that was discovered, described 10, 20 years before in a paper, in a kind of dry, abstract scientific

00:16:19.460 paper. And all of a sudden, that gene, the genetic material that can drive the growth of a normal

00:16:25.620 cell and make it malignant, make it metastatic, so metastatic, so malignant, that our best medicine,

00:16:31.920 our best minds can't stop the growth of this aberrant cell, all being driven, in part, by that

00:16:38.760 very same gene that was trapped 20 years ago in a laboratory experiment on rat cells. So all of a sudden,

00:16:45.160 this thing comes alive to you in a way that becomes consequential. If you didn't know the

00:16:50.780 identity of the gene, you would not understand why on earth this 70-year-old man in perfectly good

00:16:57.020 health is all of a sudden decimated by one or two or four or eight mutant genes in his cells that

00:17:05.240 suddenly take over and drive the growth of these cells. Anyway, that's one example that comes up over

00:17:09.740 and over again. That was the first tenet. That was the first tenet. So what was the second one?

00:17:13.100 There are many, so I'm not going to go through all the tenets. But the second one was that this book,

00:17:18.620 all these books, should be fundamentally readable by everyone. You talked about this already.

00:17:23.280 It should be like a kaleidoscope, that if you turn the book leftwards and you see it as an

00:17:29.120 oncologist, it's still interesting. The pattern changes. If you turn the book rightwards and say

00:17:34.160 you're an anthropologist, it still is interesting. If you turn the kaleidoscope upside down and shake it

00:17:41.160 and there's a new pattern that's formed, it's because you're reading the book now as a clinical

00:17:46.320 scientist. Or you're reading the book because your daughter has leukemia and you're the father

00:17:53.020 of a patient with leukemia. Or you're reading it because you yourself have been diagnosed with

00:17:58.440 breast cancer. Or your mother has. And all of a sudden, the kaleidoscope changes. But the point is

00:18:04.700 that the object remains the same. It's the same book, but you can read the book in various different

00:18:09.560 ways. You can come into it different ways. Often when I'm in, this sounds like a strange statement,

00:18:14.960 often when I come back from the wards, I reread my own writing to figure out sort of what was I

00:18:20.280 thinking then in 2008 when I wrote those sentences and how does that change now?

00:18:26.420 Has it really been 10 years?

00:18:27.920 It's been 10 years. So I should tell you that there will be a 10-year update to the emperor.

00:18:32.740 There'll be three additional chapters.

00:18:35.120 I assume one of them will be immunotherapy as an update.

00:18:37.600 That's right. So there are very broadly three sections that are updated. And I've thought about

00:18:42.740 it for a long time. So there'll be an updated section on prevention. There'll be an updated

00:18:47.380 section on early detection and an updated section on treatment. So that's the very broad three big

00:18:53.680 broad chapters. But in every chapter, there will be deeper dives into what's happened now since the

00:18:59.580 last 10 years and within the treatment section. And potentially within the prevention section,

00:19:03.840 there'll be a huge role of the immune system, which was not fully appreciated in 2010.

00:19:08.980 I thought that was the biggest distinction between the book when I read it and then the PBS special,

00:19:13.740 the Ken Burns special, which of course, again, we'll link to all of these things in the show notes

00:19:17.580 here. But one, you got to read the book. But two, I can't recommend enough the Ken Burns special.

00:19:23.280 They did an unbelievable job, I think, sharing your voice. And that was the biggest. I remember

00:19:28.720 watching it thinking, oh, wow, there's a big difference here. Because was it three installments

00:19:32.460 or five? I can't remember.

00:19:33.220 Three.

00:19:33.820 Yeah. The third installment felt like it was half immunotherapy.

00:19:36.380 It was half immunotherapy. Well, because part of the reason was that, again, this brings me to the

00:19:39.900 next tenet in the book or writing the book. The next tenet was that there is so much cancer research

00:19:45.760 in laboratories going on everywhere. And the tenet or the principle was that unless that research has

00:19:51.940 manifest itself in a human drug, in a human medicine, in a reconception of how we think about

00:19:59.740 preventing, treating, or detecting cancer, in a fundamental reconception of those ideas,

00:20:05.280 it won't get into the book. So tumor immunology, of course, had been around for a very long time,

00:20:10.240 Coley's toxins, famously, and other such efforts very, very early on. But in 2010, we were at the edge

00:20:17.340 of that moment in which we began to use tumor immunology in human beings as powerful medicines

00:20:25.700 to change the course of the disease. It was just the first trials that come out. And in fact,

00:20:30.640 the book was completed in 2009, and the first trials had not even come out then.

00:20:33.680 This is a little funny trivia story that I'd almost forgotten about this until somebody mentioned it

00:20:37.900 to me a year ago. I think the first real paper that I wrote as a fellow was on CTLA-4. It was

00:20:45.280 looking back at the series at NCI of patients who had been given CTLA-4 and responded, and this paper

00:20:51.080 basically identified the strong association between autoimmunity and their response to it.

00:20:55.920 Right.

00:20:56.100 Of course, this will be interesting later in our discussion because I want to, of course,

00:20:59.240 talk about James Allison. So we'll come back to that. But you're right. That was sort of when it

00:21:04.020 went from you had interleukin-2 that worked in maybe 10%, 15% of patients, but you couldn't predict

00:21:09.880 why. That was the bigger question. You didn't know why were some responding, and why were they only

00:21:13.780 responding with certain cancers, to where we are today, where it's really been a transformation

00:21:18.840 in the last 15 years?

00:21:20.360 Absolutely. And so when we started the film version of Emperor with Ken Burns, which I obviously was

00:21:25.980 very close, I worked very closely with Ken, I should say that the gene is also being made into

00:21:30.480 a film by Ken Burns.

00:21:31.460 Oh my God. It's like Christmas all over again.

00:21:34.140 It's almost completely shot. We're sort of moving towards editing phases and so forth.

00:21:38.560 When can people expect that?

00:21:39.700 Oh, it's already been scheduled. I think it's winter or fall 2019. So it's on the scale.

00:21:44.960 It's about a year from now.

00:21:45.820 Yeah, about a year from now. Yeah. We're working through footage and historical footage and archival

00:21:50.320 footage and so forth. But to wind backwards, the crucial piece that had been added since the book

00:21:55.780 was, of course, immunological therapy. So much of what the last episode was around this new

00:22:01.640 burgeoning field. I mean, in the first meeting that we had around the transformation of the book into

00:22:06.680 the documentary film, the first thing that was raised is what's changed? And the answer was very

00:22:11.400 obvious. What's changed is immunological therapy.

00:22:14.000 It might have been emphasized as much in the book, and I may have just missed it. Or it's possible you

00:22:18.040 also observed this as a change. But I believe it will factor into the 10th edition, which I can't

00:22:23.860 wait to get my hands on, is the role of obesity in cancer.

00:22:28.020 As if I recall, in the documentary, you said, look, this is now becoming basically the second

00:22:33.820 leading preventable cause of cancer after smoking, which was, again, I was aware of that at that

00:22:39.260 point in time because of the work that I'd been doing and sort of my little echo chamber. But I

00:22:43.920 thought, this is a PBS, this is not something that I think most people would appreciate. This is

00:22:48.760 becoming a first for many people to hear.

00:22:50.840 That's right. And the one thing that we should make clear is that every word in that PBS document

00:22:56.900 was vetted by some of the most important and thoughtful scientists and cancer biologists and

00:23:05.840 physicians and physician scientists and cancer advocates across the world. There's a backstory.

00:23:12.100 In other words, the script was vetted over and over again so that we wouldn't say things that were

00:23:17.380 misinforming the public or because this is documentaries for all time. Ken Burns's work

00:23:22.760 is evergreen. Hopefully, my work is evergreen. So it was very, very carefully vetted. And it was

00:23:28.200 quite clear by the time the documentary came out that the signals that we were picking up around

00:23:32.980 obesity and cancer were becoming extraordinarily clear for some cancers, obviously not for other

00:23:38.720 cancers. And the provocative statement that sits behind all of this is that really since the last

00:23:45.500 20-odd years, maybe even 30-odd years, we have been struggling, struggling to find preventable human

00:23:53.940 chemical carcinogens of substantial impact. Every word in that sentence is important. We have certainly

00:23:59.880 found new chemical carcinogens in humans, but often that affects small pockets of people who are exposed to

00:24:06.760 those carcinogens. We have found lots and lots of chemical carcinogens which have moderate to very small

00:24:13.320 impact if you look at populations overall. The bar might be smoking. So smoking is a good bar. This is a

00:24:20.120 smoking or tobacco smoke is a chemical carcinogen which is removable or preventable, and it has substantial

00:24:26.760 human impact across populations. Changing smoking behavior can change fundamentally the epidemiology of

00:24:32.540 cancers across nations. If that bar is smoking, we have struggled for the last 20-odd years to find things of that

00:24:39.780 magnitude and effect. The direct impact of that is when people come to me and they say, well, what do you do

00:24:46.080 to prevent cancer in yourself and your family? I have to sort of casually or not so casually admit that

00:24:52.640 not very much. I obviously don't smoke, but it's not like I'm eating goji berries or avoiding some

00:24:59.460 fundamental thing that everyone else is not in the know about because there aren't any. I mean, you know,

00:25:04.400 I'm obviously not exposing myself to these rare, unusual occupational cancer carcinogenic agents,

00:25:10.180 but I'm not doing something fundamentally changed that's different from you or anyone else to prevent

00:25:15.580 cancer. The exception to this rule of the 20-odd years of the hunt for chemical carcinogens is obesity.

00:25:24.300 Now, you and I can have a debate. Is obesity a chemical carcinogen? No, not in the traditional sense.

00:25:28.940 Is it even preventable? Maybe, but we have to think twice or three times about it. There's a role of

00:25:33.680 genetics and environment in all of this. And even do we think it's obesity per se or hyperinsulinemia

00:25:38.200 or any other endocrine? Exactly. First of all, is it an endocrine problem? Is it an inflammatory

00:25:42.360 problem? Or is it a metabolic problem? There are at least three horns. Yeah, obesity is such a

00:25:47.200 crude phenotype. That's right. To your point, has at least three, if not six underlying phenotypes

00:25:54.820 that each of which mechanistically would make a lot of sense for accelerating cancer.

00:25:59.320 This is an immunological phenomenon. The cytokines, the inflammatory, yeah, all of these things.

00:26:03.160 Going back to your general point that obesity was becoming identified more and more clearly

00:26:07.580 as one of the potential causes of cancer, or I should say strongly correlated with the

00:26:13.500 development of cancer so strongly that we think that there's a causal link based on all everything

00:26:18.320 that we know about epidemiology, some cancers, that we began to take this seriously. And we're

00:26:22.960 still taking it seriously. But as you're pointing out, there are several horns underneath that blanket

00:26:27.680 of obesity that we understand very crudely. And we have to figure out which of these is driving

00:26:31.780 the cancer risk. I would like to spend the next six hours discussing with you the tenets of writing

00:26:36.980 for selfish reasons. But instead, I'll punt that to a... We'll have dinner in a couple of weeks and

00:26:41.780 we'll finish that discussion. I want to talk about another book you wrote that doesn't get as much

00:26:45.060 attention, which is The Laws of Medicine. You wrote that after Emperor Before the Gene, correct?

00:26:49.720 That's right. So The Laws of Medicine has a very different mandate, as it were. And that's because

00:26:54.580 the book came out in association with TED. They had commissioned 10 books by 10 thinkers around the

00:27:01.180 world. And they asked me to write a book on that. And it's necessarily a small book. It's really a,

00:27:06.440 you know, the mandate was to write basically a 75-page book expanding on a single very,

00:27:11.780 very incisive idea. So that's The Laws of Medicine, yes.

00:27:14.120 If I got them correctly, the three laws are a strong intuition is much more powerful than a weak test.

00:27:19.200 How did you think of that? And what is the most important application of that law to the way you

00:27:24.900 think about medicine or specifically oncology today?

00:27:27.500 This, to me, is one of the great neglected ideas in medicine, perhaps one of the great

00:27:31.640 neglected ideas in the world. This idea initially comes from Thomas Bayes. This is a Bayesian idea.

00:27:37.740 Thomas Bayes was a cleric, but by evening he was a mathematician and an economist. And his work

00:27:45.260 leads to one of the most seminal and funny thought experiments that I've ever encountered,

00:27:48.480 which is the following. And I sometimes quiz my daughters with it, which is the following. This

00:27:53.200 is not Thomas Bayes' own example, but it arises out of Thomas Bayes' work. And he, one might imagine

00:27:58.920 going to a street fair and encountering a man who's tossing coins. And he tosses coins and your job is

00:28:06.760 to predict whether the next coin flip is going to be heads or tails. And so he tosses the coin 20 times

00:28:13.160 and all 20 times its tails. So then he turns to the crowd and he says, what's the next coin flip

00:28:20.760 going to be heads or tails? Now, the mathematician in the crowd, who's the professor of mathematics...

00:28:27.220 Says 50%. 0.76

00:28:28.220 Says 50%. And, you know, everyone says, absolutely right. But the child in the crowd says, no, no, you don't

00:28:34.980 understand. This is a stupid problem. It's the coin's rigged. The coin has only, it has two heads or two tails, 1.00

00:28:40.940 as the case may be. And the child's right. And what's important about that insight is that the

00:28:45.940 mathematician imagines the world, this in this case, this is not a stab at mathematicians in general,

00:28:51.000 but the professor of mathematics thinks of the world as having no history, as having no a priori's.

00:28:57.640 It's a world that's created de novo every time. The coin is flipped and its heads and tails equal

00:29:02.420 every time. But the child knows, and all humans know, that in fact, the world doesn't behave like

00:29:09.180 that. Everything has priors. And you need to understand those priors before you can understand

00:29:14.380 the posteriors. There's wisdom in that idea. And it took someone like Thomas Bayes to figure that out,

00:29:19.460 that most of our lives, we aren't living our lives like the crazy mathematician professor.

00:29:25.160 We are living our lives like the child. We're thinking to ourselves, what was the prior antecedent?

00:29:29.800 Imagine this is true for any corner of your life. The first question you ask yourself when you're

00:29:35.040 trying to solve a problem, trying to understand the cosmos, trying to understand something,

00:29:39.760 you ask yourself, what was the prior like? Did the sun set in the west last night? And how about the

00:29:45.080 night before? And maybe I don't need to create a formula to figure out whether the sun is going to

00:29:50.240 set on the west or the east tomorrow. It's because it's set on the west every time. There are obviously

00:29:54.720 loopholes and gaps to this kind of thinking. There are surprises that you can miss. So

00:29:58.500 Bayes' fundamental idea was that you can only interpret a test in the light of what that test

00:30:05.100 is predicted in the past. It's an extraordinarily important idea in the way we think about the universe,

00:30:10.600 that the past performance of a test tells you something, not everything, but tells you something

00:30:16.280 about the future performance of a test. And you can apply it to many, many things in the world. You

00:30:21.300 can apply it to any kind of thinking, economic thinking, climate change-oriented thinking, that

00:30:26.500 the past is a guide to the future, not only in a kind of loose way, but you're really using a real

00:30:33.240 stat weighted strongly by the past. And this, of course, applies to medicine. And it's a forgotten rule in

00:30:39.420 medicine. Although it does seem like one of the things that I've always felt physicians innately

00:30:43.920 do well without realizing it, which is the opposite of where I want to be, not to take the pot shots,

00:30:50.160 but I think where we do very poorly is in understanding asymmetric risk. So Nassim Taleb has

00:30:56.720 written a lot about asymmetric risk. And I think he's absolutely right to be critical of not just

00:31:01.580 physicians, but basically most people. So my argument is we are innately wired to be Bayesian.

00:31:07.340 Again, we are absolutely not innately wired to appreciate risk.

00:31:11.260 And both of these are important.

00:31:12.840 That's right. So one of them we have to hone so much more because I think it just doesn't

00:31:16.900 come naturally. I played a funny experiment, which you'll appreciate. I'll send you the

00:31:20.720 list and you'll have a field day with it. It's 20 questions and each one is a quantitative

00:31:26.100 question with an answer, but they're not obvious. You would never, it's not like how many,

00:31:30.140 you know, presidents were there or something where you might know the answer.

00:31:32.880 You ask the group, we're going to give you these 20 questions. I want you to answer each

00:31:37.900 question, not with a number, but with a 95% confidence interval.

00:31:41.560 You know the game?

00:31:42.240 No, I don't know the game. It's an interesting game.

00:31:43.700 Okay. So an interesting game. So at the end of 20 questions, if you've done it correctly,

00:31:47.080 you should have 19 out of 20 of those ranges. Correct. I've never met a person who can come 0.94

00:31:51.980 close. You almost without exception get like seven of the 20, right? You can't even contemplate

00:31:59.400 what that variability is. So the second law was that normals teach us rules, outliers teach us

00:32:06.720 laws. So this, of course, these were very carefully, I mean, I thought I've spent a lot of time

00:32:10.880 thinking about them. Of course, this is the anti-Bayesian law. This is the exactly what

00:32:14.220 you're talking about. This is the idea that simultaneously in the medical brain has to live

00:32:18.760 the idea that the Bayesian idea that when you hear hooves think horses, not zebras, famous medical

00:32:24.420 tenant. Hooves beats outside your window are likely to be horses. They're very unlikely

00:32:29.220 to be zebras. The second law is once in a while, they are zebras and you need tools. You need special

00:32:34.780 ideas, special tools to figure out what these outliers look like, who they are, how to find them,

00:32:41.660 and how to quickly find them so that you can identify them and triage them differently. So in some

00:32:48.000 ways, these two laws are yin-yang. They polarize against each other. And what's interesting about them

00:32:53.960 is that they both in medicine can both can be simultaneously true in the same way as our

00:32:59.080 assumption of a symmetric risk is simultaneously true with the idea that our understanding of the

00:33:05.240 of a world in the Bayesian way is helpful and important. So the second law is about how do you

00:33:10.400 identify outliers, how outliers tell us about the nature of normalcy, how they tell us about how

00:33:18.260 complex interactions can produce occasional far outsiders and how those outsiders really challenge

00:33:25.860 us to define what these interactions look like in real life. I mean, the simple example is, it's very,

00:33:31.100 very hard to figure out the genetics of any disease without finding the rare people who have the disease

00:33:39.480 as a consequence of a mutation. The classic example, of course, is that we would not have an idea

00:33:44.320 how to regulate our body or to prevent heart attacks. If we hadn't paid extraordinarily close

00:33:51.380 attention to a small family, small groups of families that had a mutation in the gene that controls

00:33:58.440 cholesterol metabolism, most people don't have this mutation because those families are quickly

00:34:02.780 extinguished, we think, because they die of heart attacks. They have all sorts of problems, they die of heart

00:34:07.480 attacks. But by paying extreme attention to this one family that has a mutation, they're rare,

00:34:12.700 scientists all of a sudden uncovered a whole cosmos in which we understand now cholesterol metabolism.

00:34:20.200 And because of that cholesterol metabolism, other scientists eventually developed the first statins.

00:34:24.720 So all of this, you know, the fact that hundreds of thousands of people in the world are taking

00:34:28.980 this medicine to prevent heart attacks, tracks back to a rare family where, because of a genetic

00:34:35.420 mutation, they had a very, very high risk of cardiac disease.

00:34:38.720 And of course, now we know there's a whole family of those people. There's at least 2,000 of those

00:34:42.860 mutations that produce that phenotype. And these natural experiments are actually a remarkable thing.

00:34:48.380 And of course, bringing it back to the gene, to now have a tool, a probe, to be able to explore that

00:34:54.360 is amazing. The third law, for every perfect or exceptional medical experiment, there is a human

00:35:01.420 bias that goes along with it.

00:35:02.920 Again, these are unique to medicine. And what's interesting about them is that they are unique,

00:35:06.840 to me, at least, though it's interesting, is that they're unique to the day-to-day practice of

00:35:10.860 medicine, but they apply for every, I think, they apply to every corner of life. You don't have to

00:35:16.560 be a doctor to realize that for every time we think of something, if you're really skeptical thinkers,

00:35:22.640 we have to think about the bias that comes intrinsically with that thought. In other words,

00:35:26.780 every single declarative claim about the universe that we're making must have necessarily a

00:35:34.620 declarative bias that comes with it. They're matched. If you want to be a scientist, if you

00:35:38.840 want to be a skeptic, your real job, aside from being creative and designing experiments,

00:35:44.720 is to find for every single declarative claim that you're making, the bias that drives sitting

00:35:51.400 like a little devil buried inside that declarative claim. Because I promise you, everyone has one.

00:35:57.020 Each one of these claims has one.

00:35:58.340 And Richard Feynman, who listeners of this podcast know, is one of my heroes. One of my

00:36:02.560 kids is named after him.

00:36:03.900 Is he named Feynman or Richard?

00:36:05.020 His middle name is Feynman.

00:36:06.080 I see.

00:36:06.380 Yeah. And my wife at first was like, why would we do that? But then once she had read,

00:36:10.800 surely you're joking, Mr. Feynman a couple of times, because she'd read it once before she

00:36:14.100 realized why. But Feynman said eloquently, right? The job in science is not to fool yourself,

00:36:19.120 and you're the easiest person to fool.

00:36:20.700 That's right.

00:36:21.180 So to your point, the only thing I would disagree with what you've said, Sid, is I don't think these

00:36:24.500 are unique to medicine. I think these are laws of science, and maybe laws of life on some dimension.

00:36:30.700 So the mandate here really was to use this book as a kind of springboard to challenge the way we

00:36:36.640 think about virtually all aspects of how we live. You're going to read an article in the New York

00:36:42.900 Times tomorrow that will make some claims about some politician somewhere in Oklahoma, or you're

00:36:50.460 going to read about an economic paper that is now being presented to the feds. And your job is to

00:36:55.920 be adequately skeptical about it and understand what are the priors? How do we explain this particular

00:37:02.100 fact based on the priors? Do the priors matter? To use a very topical example, does someone's past

00:37:08.480 behavior in college or in school, high school, tell us about how he or she is going to be a judicial

00:37:14.680 candidate?

00:37:15.140 Right. Or does their behavior under scrutiny in any way predict their behavior when no one's looking?

00:37:20.880 That's right. That's another question. The second question that you might ask here is,

00:37:24.960 again, it's an outlier question, right? Scrutiny versus not scrutiny. Do they become outliers to

00:37:30.200 themselves if they're scrutinized versus when they're not scrutinized? And the third question is,

00:37:35.080 when I read this story, what are my biases? Am I reading it because I'm a man? Am I reading it because

00:37:40.120 I have a particular experience in my own lifetime, myself, my sister, my daughter, my friend, my

00:37:47.720 colleague? Is that coloring the way that I read this particular story? So absolutely. And this

00:37:53.780 thinking goes on over and over again. These are circular processes. Do those biases have priors? Do

00:37:59.320 the priors have biases? And so forth. So you can use these. I mean, I certainly use these tools in

00:38:04.580 clinical practice.

00:38:05.780 Yeah. It's a beautiful, as you said, it's a very short book, but again, we'll certainly make sure

00:38:09.900 people have this link to it because I think it's a great way of teaching people how to think. And

00:38:15.040 that's, again, I just don't think we're innately wired for every facet of thinking. Now, going back

00:38:20.100 to your background in immunology, very recently, we had a very exciting, for those of us in this

00:38:25.260 space, very exciting Nobel Prize awarded. And it was, as we were discussing before we started

00:38:30.120 recording, it wasn't so much around immunotherapy, but a very specific element of it, which are the

00:38:34.920 development of checkpoint inhibitors. Two of them in particular were basically acknowledged

00:38:39.360 here, CTLA-4, which I mentioned earlier, and PD-1. Tell me why these are so exciting.

00:38:47.180 They're exciting for many reasons. They are exciting for, again, some history and some background. The

00:38:52.620 story, as it were, is exciting because if you ask the question in the 1990s, what's the relationship

00:38:59.780 between cancer and the immune system, you'd get a kind of diffuse answer. You'd get an

00:39:04.420 answer which could be very unclear because there were all sorts of complicated lines of

00:39:08.820 evidence. One line of evidence was that in patients with a complete collapse of the immune

00:39:13.780 system, such as patients with HIV, or complete collapse of at least one wing of the immune system.

00:39:20.500 They would get these cancers nobody else could get.

00:39:22.640 But what's interesting about them, and this is where the thought experiment, this is where the

00:39:25.780 brain cells start sort of ticking and wondering, they would typically get viral cancers, viruses that

00:39:31.880 would now get unleashed. We now know many of them. Viruses like human papillomavirus, they would get

00:39:37.700 cervical cancers, they would get anal cancers. So a lot of viral cancers, but patients with HIV did not

00:39:44.780 generally get lung cancer.

00:39:47.120 I never even thought of that, Sid. So we have enough data to know that they either at no greater

00:39:52.740 prevalence or even at a lower prevalence when you start to talk about, you know, for example,

00:39:56.780 lung cancer or GBM or pancreatic cancer. I never thought of that.

00:40:00.740 Yeah, it's very important. So the data are mixed because, of course, in the 1980s, they weren't

00:40:05.700 living long enough. So all that we know is that the first groups of cancers that cropped up in these

00:40:11.000 men and women were not lung cancers. They were not pancreatic cancer. So at the first pass with that

00:40:17.840 cut short data set, you might begin to imagine then if that's the case, the immune system completely

00:40:23.200 collapses, you only get a certain kind of cancer, then what is the possible role of the immune system

00:40:29.420 in cancer control? And if you were a nihilist, you might have given up in the 1990s and said,

00:40:35.300 well, you know, take the whole immune system away and nothing really happens even with this cut short

00:40:40.320 data set. So maybe there isn't such a complex interaction. Well, it turns out that people like

00:40:45.300 William Allison and his Japanese colleague did not put the immune system away. They put that data

00:40:51.900 aside, said, this goes back to the laws of medicine. They said, sure, it tells you a little

00:40:56.340 bit, but it doesn't tell you the full story. The full story turns out, there's one layer deeper.

00:41:01.900 The full story turns out that in people who don't have a collapse of the immune system, whose immune

00:41:06.420 system is otherwise intact, cancer cells, not all cancers, but some cancers make specific factors.

00:41:14.640 In fact, they evolve. The word make is the wrong word here. They evolve so that they put up specific

00:41:20.720 factors, put up specific signals that inactivate the immune system or that make the immune system

00:41:27.160 no longer able to kill or recognize the cancer cell. And the identification of these specific

00:41:33.680 pathways, these specific factors was what led to the Nobel Prize because in further work,

00:41:39.180 Jim Allison and again, his colleagues, this is a big wide field, showed that if you inactivate these

00:41:47.680 specific factors, if you drive nails through them using a variety of methods, then all of a sudden,

00:41:53.180 the cancers become revisible to the immune system and the immune system can attack and kill the cancer

00:41:58.760 cells. Now, why is it exciting? It's exciting for many reasons. First of all, it's important to realize

00:42:03.240 that not all cancers respond. We don't know why some do and some don't. Melanoma is highly responsive.

00:42:08.700 It's an immunologically engaged tumor. Is it because the melanoma has so many antigens that it will

00:42:13.840 suddenly be more recognizable? Is it because the skin is such a lymphoid organ? Is it the right

00:42:18.340 environment? We don't know all of these questions. The other thing to realize is that people have

00:42:22.960 responses and some people continue to respond. Some other people, sadly, will relapse and they'll go,

00:42:28.540 you know, the cancers will start growing back in the context of their re-educated immune system.

00:42:34.200 And that leads to the second question, which is what happens then? Is there a second pathway?

00:42:39.540 And this is an important idea, I think, that goes back to your first question, which is about whether

00:42:43.700 this is depressing or not. The important thing is once you drive a single stake through cancer's heart,

00:42:51.020 it's like placing the first crampon on a climb. You see, if there's no crampon that's placed on the

00:42:57.000 climb, you can't climb a mountain. It seems like a wall. It's a blank wall. And you don't know where

00:43:02.460 to go left or whether to go right or what to do. Once you plant that first crampon...

00:43:07.340 And it sticks.

00:43:08.120 And it sticks. You, all of a sudden, the whole face of the mountain becomes...

00:43:12.660 It's a different mountain now.

00:43:13.360 It's a different mountain now. Because now you can ask the question,

00:43:15.860 because the first crampon was planted in that particular place...

00:43:19.040 And it stuck.

00:43:19.900 And it stuck.

00:43:20.440 I have a new vantage point.

00:43:21.840 Exactly.

00:43:22.100 The perspective has changed.

00:43:23.160 That's right. And so this is the important piece to realize about cancer research and

00:43:27.240 perhaps about all research, is that the first crampon or the first stake through cancer's

00:43:31.920 heart is an incredibly important stake because then you can ask what I call linear questions.

00:43:37.020 Before that first one is placed, the questions are nonlinear. You don't know where to place

00:43:40.880 it. The whole map is open. Once you drive a stake through the first question, the world becomes

00:43:47.340 more linear. You can now ask the question, what's the mechanism of resistance to that?

00:43:51.060 And when you find that, what's the mechanism of resistance to that?

00:43:55.200 And so all of a sudden, the perspective is different because you've climbed through that

00:43:59.040 first. And that's usually where the Nobel Prizes are given. The Nobel Prizes are given often

00:44:04.200 for planting the first stake through the heart of a disease, through the heart of a problem.

00:44:09.080 And that's why this Nobel Prize is important. It's not, of course, as you're saying,

00:44:12.260 this does not encapsulate the field of tumor immunology in general. It's a wide field.

00:44:16.280 Our own lab does a whole bunch of work in tumor immunology, but we're on the shoulders of those

00:44:21.280 prior giants, as it were. And so that's the recognition that's been given here.

00:44:24.880 You said something that I think is so important and worth reiterating. There's also something about

00:44:29.240 immunology and immunotherapy that's quite interesting as far as planting that stake,

00:44:33.300 which is the durability of response. It is often the case that when you have an immunologic

00:44:37.340 remission, it is a durable remission. And it's not always the case, but it is much more likely the

00:44:41.740 case than when you say, for example, see a chemotherapeutic remission or even a surgical

00:44:46.060 remission. On a personal level, I have a very close friend who was diagnosed with colon cancer

00:44:51.880 10 years ago at a very young age. He might've been 40. And that was unusual in and of itself.

00:44:58.760 I remember talking. I went to the hospital when he had his colectomy, spoke with the surgeon after.

00:45:04.360 It sounded like a horrible case, huge tumor. The mesentery was full of nodes, which I assumed would be

00:45:10.520 positive. Every 26 nodes sampled all came back negative. My friend was adopted. We later realized

00:45:16.640 he had Lynch syndrome. Eight years later, he goes on to develop pancreatic adenocarcinoma,

00:45:21.540 unresectable. So it's encased the mesenteric vessels. He cannot have a Whipple procedure.

00:45:27.420 And as you know, and unfortunately many people listening to this will know, that is a death

00:45:31.980 sentence. It's a non-negotiable, you will not be alive in nine months. And he was put on an anti-PD-1

00:45:38.860 therapy because he happened to have, as you would know, patients with Lynch are going to be much

00:45:43.000 more likely to be susceptible to checkpoint inhibitors. Two years later, he's disease-free.

00:45:47.500 Just unbelievable stories. And I think your point is an elegant one that I'd never really

00:45:52.320 thought of before, which is focus less on the fact that we haven't solved the problem and more on the

00:45:57.000 fact that we have made a finite and real step towards establishing a new place, a new location

00:46:03.400 for which to view this disease.

00:46:04.620 That's right. So I describe this as taking a non-linear problem into a linear problem. Now,

00:46:10.000 of course, the answers are often not linear per se, but it gives you a route. I mean, we'll come back

00:46:14.880 to the metabolism study that we did with Lou Cantley. That is a great example of taking a non-linear

00:46:20.340 problem and converting into a linear problem. Now, I'll tell you about that in a second. But

00:46:24.660 anytime, this was the case with Gleevec too.

00:46:27.300 Let's tell people what Gleevec is. Yeah, it's a good story.

00:46:29.520 Gleevec is a good example of a drug where we learned to target a mutant cancer gene,

00:46:36.060 and it's actually the gene product. So in that case, the cancer, it's a blood cancer called

00:46:40.900 chronic myelogenous leukemia, was also a death sentence.

00:46:43.600 And also GI stromal tumors.

00:46:45.060 And GI stromal tumors, yeah. Chronic myelogenous leukemia was a death sentence. People had to

00:46:49.580 go through transplants. I've watched probably a dozen in the pre-Gleevec era, a dozen patients die

00:46:55.980 of chronic myelogenous leukemia or the complications of transplant. I trained briefly as a transplanter.

00:47:01.320 And then all of a sudden, through the work of many people, including Brian Drucker,

00:47:05.560 a drug was discovered. So before that, scientists figured out, based on very careful genetic analysis,

00:47:12.220 that the tumor was being driven by really the work of one gene. And the gene was called BCR-ABLE.

00:47:18.580 It's an oncogene. It's not found in normal cells, but found in these cancer cells.

00:47:22.680 And the product of that gene became like an engine, like a manic engine that was driving

00:47:28.840 these blood cells to go crazy and make more blood cells and proliferate and proliferate and

00:47:34.500 proliferate. And it was this engine gone wild inside a cell. And every time the cell asked the

00:47:39.140 question, should I divide, rather than looking at its normal state or nutrients, metabolic state,

00:47:44.280 et cetera, the only answer it would ever get was from this engine saying, yes, go ahead, make another

00:47:48.220 cell divide, et cetera, et cetera, et cetera. So this was identified, that the gene product was

00:47:54.080 identified, the protein product of the gene was identified. And then through an elaborate series

00:47:58.540 of experiments and circumstances, chemists found a way to actually jam the engine. One molecule,

00:48:04.860 human beings, we stitch this molecule together. It's a remarkable achievement. And once the engine

00:48:09.820 is jammed, it suddenly turns out that the disease goes into remission. This was one of the first

00:48:14.040 examples, there are others, of so-called targeted therapy, where you synthesize a chemical to jam

00:48:19.200 cancer's engines. In a cancer-specific way, it doesn't affect normal cells. But what's interesting

00:48:24.500 about that is that some people develop resistance because the engine, cancer cells evolve, and they find

00:48:31.960 a way of resisting. But that resistance, it becomes a linear problem. You figure out what the mechanism

00:48:37.540 of the resistance is, and you drive a second stake through cancer's heart. And when it becomes

00:48:41.660 resistant to that, you drive a third stake and so forth. So all of a sudden, the problem, which seemed

00:48:45.720 like a big blank rock, became a linear problem. So that goes back to another example of how that first

00:48:51.420 stake or the first crampon really helps with the problem. And now for immunotherapy, immunological

00:48:56.080 therapy, the roadmap is much clearer. Why do some tumors respond and why some don't respond? Is it a question

00:49:03.680 of the environment that the tumor is sitting in? Is it blood vessels? Is it tumor cells? Is it immune cells?

00:49:08.780 These are answerable questions. These are so-called linear questions. But the first big step here was

00:49:14.120 to define the problem. This is really a nice step off to exactly where I know we want to go, which is

00:49:19.000 we've got surgical oncologists, medical oncologists, radiation oncologists. In many ways, a subset now of

00:49:25.720 medical oncology is immunobased oncology. The work that you, Lou, and many others are now doing is

00:49:31.960 potentially another branch of oncology called metabolic oncology. So you're beyond gracious

00:49:39.160 in your suggesting I have even something to do with helping that evening turn into a, what sounds

00:49:44.740 like a very productive collaboration between you and Lou. And I'll be talking with Lou as well

00:49:48.240 in the next month, I'm sure. But let's talk a little bit about what came out of that collaboration

00:49:54.180 and certainly bring it back to Ben Hopkins' paper that was in nature, I believe.

00:49:59.220 It was in nature, yeah. Two weeks, three weeks ago, yeah.

00:50:00.820 Yeah. So let's talk a little bit about the question. What were you trying to understand?

00:50:04.840 So the question is, there's a very wide question and then there's a very narrow question.

00:50:09.680 The wide question is how does the body's metabolic state affect cancer? It's a very big question because

00:50:18.300 cancers, like all cells, are eating nutrients as well in order to grow. And the question,

00:50:23.300 therefore, is the cancer eating a different set of nutrients than normal cells? This work goes back

00:50:28.340 to famous work by Otto Warburg done in the early 1900s, 1920s, is that right? Yeah. Where Otto Warburg

00:50:36.220 was one of the first people to make the hypothesis that there's something fundamentally different

00:50:40.480 about, I mean, we won't go into great details, but it's fundamentally different in which the cancer

00:50:44.700 cells metabolize compared to normal cells. They use fundamentally different pathways to metabolize.

00:50:50.500 And if you could find a way to target these metabolic alterations in cancer cells, you'd find

00:50:55.240 an anti-cancer drug. So, but that question has been hanging around our field for a long time. And as

00:51:01.100 our understanding of normal metabolism has changed, we've begun to identify not just one, but dozens

00:51:07.380 of nutrient pathways that cancers use that may or may not be different from normal cells, maybe slightly

00:51:13.460 different from normal cells, maybe a lot different from normal cells. So that's one big question.

00:51:17.820 The second question, which is a slightly narrower question, is that when you give a drug, whatever

00:51:24.820 drug it might be, your favorite chemotherapy, your favorite drug, any drug, does it change

00:51:30.300 the metabolism of the cancer cell? And does it change the metabolism of the body? And could

00:51:36.140 this be a mechanism by which cancer cells become sensitive or resistant to chemotherapy? So these

00:51:40.600 are two related questions. Again, to reiterate, one question is, how is the cancer cell's metabolism

00:51:46.040 different from the normal cell in normal circumstances? And the second question is, how is a cancer cell's

00:51:52.160 metabolism deferred from a normal cell in the context of giving a drug?

00:51:56.040 Yeah. In other words, can a nutritional state be exploited and or a drug sensitivity be exploited

00:52:00.240 through a nutritional paradigm shift?

00:52:01.040 Exactly right. Those are the two questions. So we focused in this particular study, we actually

00:52:04.680 are interested in both. I'm interested in both. But we focused on the second question. And the second

00:52:09.780 question in this particular case was that there was a very promising group of medicines

00:52:13.400 that was being used in clinic. In fact, I had used them as a trialist. They had come really out of

00:52:19.620 Lou Cantley's path-defining work. He had defined the pathway or the signals that these medicines

00:52:25.520 attack. And there was an enormous amount of optimism because this was considered a fundamental pathway.

00:52:31.520 This is the beauty, by the way, of being in a place like New York or Boston. It's, you can talk

00:52:35.720 about someone doing this. I mean, Lou was in Boston when much of this work is done. He's now here in New

00:52:40.440 York. And you don't have to collaborate with a guy across the world. You can collaborate with

00:52:44.160 the guys on the Upper East Side instead of Chelsea.

00:52:46.580 That's right. So Lou's work over the last decades had been to define this pathway and ultimately led

00:52:53.040 to the formation or the creation of these new medicines. They're called PI3 kinase inhibitors,

00:52:57.660 and they go by fancy names like duvelacib and things like that. But anyway, when they came to clinic,

00:53:04.160 surprisingly, there were some responses. But there was a lot of resistance in patients that tumors

00:53:09.460 were resistant to the drugs and didn't respond. And that was a puzzle that Lou had come up with

00:53:13.960 in his own work. And then separately, seven miles uptown, I was scratching my head about the same

00:53:20.420 puzzle because I'd been using these same drugs in cancer patients and finding that patients became

00:53:26.480 resistant or were resistant to start. And so what was chalked out on a napkin that evening was we had

00:53:35.580 thought of lots of ways that these patients could become resistant. We had thought, oh,

00:53:39.360 maybe the tumor had a mutation.

00:53:41.180 Well, and there's one other point, I guess, to add just that we remember, but maybe it's worth

00:53:45.080 reiterating, is a lot of the patients that were on these began to develop phenotype. They looked

00:53:50.820 like they were diabetic.

00:53:51.740 Yeah. So that's the point that I was coming to in a second.

00:53:53.560 We didn't know if that meant that that was the source of the resistance.

00:53:56.160 That's right.

00:53:56.600 But was it true, true, and unrelated?

00:53:57.800 Yeah. The question was, was it true, true, and unrelated? I mean, you can think of many other

00:54:00.940 mechanisms of resistance. You can say the tumors got a mutation. You could say the host had some

00:54:05.900 problem. You could write down on the piece of paper a thousand ways. Those were the traditional

00:54:09.820 ways that one would explain resistance. The traditional ways of thinking about tumor resistance

00:54:13.480 is mutation. The host eats up the drug. The non-traditional way, and this is what the innovation

00:54:19.300 in the study was, that what if this diabetes that we were observing, the high levels of sugar and the

00:54:25.240 high levels of insulin, insulin being the most important piece of this, what if the drug was causing

00:54:30.220 diabetes separately from the tumor, given to normal people that the drug would cause diabetes,

00:54:35.600 in some people worse than others. And that diabetic phenotype, that diabetic state,

00:54:40.340 the hyperinsulinemic state, was being used or exploited by the tumor to essentially become

00:54:46.820 resistant to the drug. It's a little bit like, and the analogy that I drew, I remember,

00:54:51.200 on the napkin.

00:54:52.100 It's funny. I still remember the table we sat.

00:54:54.320 Yeah, I still remember the table. Yeah, exactly right. The drawing, so Lou and I were batting this

00:54:58.400 idea back and forth while you were eating nothing. And the idea was that, to me, it reminded me of

00:55:06.280 the famous story of the woodcutter who's sitting on a tree limb and chops off his own limb and falls

00:55:10.960 down. And I drew this picture, I remember, of the woodcutter on a tree limb because what happens is

00:55:16.140 that the body mounts an insulinemic response to the drug. That insulinemic response goes to the cancer

00:55:23.780 and starts feeding the cancer. And then the cancer becomes resistant. So you essentially undo all the

00:55:29.120 good that you've done with the drug by cutting off your own limb because of this intrinsic circuit.

00:55:34.600 And the long and short of it is that that's basically, in animal models, that happens to be

00:55:38.480 true. So we showed it using formal systems and formal methods that this insulinemia is a consequence

00:55:43.940 of the drug, has nothing to do with the tumor, has nothing to do with anything else. If you give the drug,

00:55:47.360 the drug goes into the liver and the pancreas and causes a sort of a pre-diabetic state that's often

00:55:53.240 worsened if you're already in a pre-diabetic state. This hyperinsulinemia is used then by the tumor

00:55:59.120 to become resistant even while the drug is present. It becomes a pathway by which the tumor becomes

00:56:04.760 resistant. And most interestingly, that you can paralyze this resistance by putting animals on a

00:56:11.420 ketogenic diet. So basically, there's never any sugar source to drive the insulin. You blunt the

00:56:17.120 insulin response so acutely that you can no longer get the insulin high. Again, this is not to be

00:56:22.420 confused with the idea that this is a sugar-feeding tumor idea. This is not a sugar-feeding tumor

00:56:27.840 paradigm. Because to be clear, on a ketogenic diet, your glucose levels might go down from normal,

00:56:34.520 but it's not an enormous reduction. Even in a complete fast, you'll still maintain at least

00:56:39.700 three millimolar of glucose. It's the insulin that becomes virtually unmeasurable.

00:56:43.760 That's right. So this is an insulin feeding the tumor question. And so I think the three points

00:56:48.900 that need to be made about the study to be very clear. First of all, it's an animal study. We're

00:56:53.320 now launching a human study that will launch in November with patients with lymphomas,

00:56:57.720 endometrial cancer, and breast cancer, particularly triple negative breast cancer I'm very keen on

00:57:02.300 studying. Because this is a cancer that has so few options. That's right. There are very,

00:57:05.340 few options. So that's the human study. The second point that is worthwhile making is that it's sort

00:57:10.680 of like, folks, don't try this at home. This is a very particular study with a very particular drug

00:57:16.080 on cancers combined with a diet. The study only worked when the drug and the diet were combined.

00:57:22.360 It does not mean that the ketogenic diet is going to prevent cancer. We don't know this. It's an open

00:57:26.760 question. We're actually studying that separately.

00:57:28.500 Yeah. And I want to talk just a little bit about that paper because that's, boy, I sure want everyone to

00:57:32.480 make sure they hear that loud and clear. Because few things upset me more than when I spend a little

00:57:37.000 too much time on Twitter and I come across people who seem to want to claim that if you're on a

00:57:40.880 ketogenic diet, you can't get cancer. Or if you have cancer, just go on a ketogenic diet. You're

00:57:44.980 going to be fine.

00:57:45.320 That's not, in fact, to be very clear.

00:57:47.200 Well, the study, in fact, demonstrated that that was not the case.

00:57:49.440 Not only that, the study also demonstrates that some cancer models, including leukemia,

00:57:53.360 is accelerated on the keto diet. In fact, we have a big follow-up study to try to figure out why

00:57:59.640 some cancers are accelerated by keto alone. But when you combine it with the drug,

00:58:03.560 they actually go back down into deep remission. The third point to make is that it seems to be

00:58:08.860 mutation agnostic. And by that, I mean, this is a very important idea, which is that 12 tumor

00:58:14.960 models responded. 12.

00:58:16.500 I was not aware that it was that high.

00:58:19.340 Yeah, 12. And actually, all 12 that we tested responded. Every one of them responded. And it didn't

00:58:25.440 matter what oncogenes you had or what tumor suppressor genes were mutated. They all responded.

00:58:30.020 Two different levels. Leukemias respond, then they relapse, at least in the models. Endometrial 0.58

00:58:36.200 cancers, we can't get them to grow. Pancreatic cancers respond extremely strongly. So it's another

00:58:40.860 example where you're not targeting, it seems, a mutation. This is not like Gleevec, in which you're

00:58:46.300 driving a stake into the engine of the cancer, or that engine.

00:58:49.940 Yeah, this is a much more global assault.

00:58:51.640 Yeah, it's a much more global assault. In fact, in some ways, it's parallel or akin to this

00:58:57.420 immunotherapy in the sense that the immune system doesn't care if you have RAS or doesn't seem to

00:59:02.400 care at the first approximation, whether you have a mutation in one gene or another gene, RAS or not

00:59:06.960 RAS. It will kill the tumor cells based on its characteristics of what it sees in the tumor.

00:59:12.020 Similarly here, the metabolism seems to be tumor-wide, but not single kinds of cancer-specific.

00:59:18.240 We don't know this to be the case. The 13th model that we try, maybe this was the one that

00:59:22.600 won't work. But this has generated a lot of excitement for this reason.

00:59:26.160 You know, I have a friend who has definitely HER2-new positive, maybe ER positive, PR negative,

00:59:31.180 but has stage four breast cancer, is in a PF3K inhibitor trial out of Dana-Farber,

00:59:35.900 has been on the trial for probably five years. She's the only survivor. Here's the interesting

00:59:39.680 thing. She's been on a ketogenic low-carbohydrate diet the entire time. 0.77

00:59:43.680 While she was taking inhibitor?

00:59:44.940 Yes.

00:59:45.300 Okay.

00:59:45.720 She's still on the inhibitor. 0.85

00:59:47.220 And so what's interesting is I've told Lou about her case because she's fastidious in this. And

00:59:53.800 she's able to get her hemoglobin A1c now under six, which for many of those patients is very

00:59:59.320 difficult.

00:59:59.720 Very hard to do, yeah.

01:00:00.640 When you, me, and Sid had lunch once about two years ago, this is the very first time you ever

01:00:05.700 showed me the data.

01:00:06.980 Yeah, that's right. We brought a computer.

01:00:08.160 We were at that Indian place or that Chinese place on 2nd Avenue.

01:00:11.360 That's right.

01:00:11.680 Yeah. And when you showed me those data, I just immediately thought of her.

01:00:15.500 So it's actually been on my list to introduce Lou because she doesn't live in New York,

01:00:19.760 but I want to bring her out and I want you guys to meet her because

01:00:23.020 the combination of this PI3K inhibitor and this dietary choice she's made,

01:00:27.940 going back to your second rule, right? She's the outlier that is giving you something to probe. 0.54

01:00:32.720 Yeah. We hopefully will have many such patients. I don't know. You know,

01:00:35.500 this is why we're running the study.

01:00:36.840 Sid, as you can probably imagine, I could spend another 12 hours speaking with you,

01:00:40.820 but I want to be respectful of your time. I know we're a little late in the day. We got a

01:00:44.260 later start than we wanted and you've got a long trip home and I want you to get to see your

01:00:48.040 family. The last question I want to ask you really is one that I need you to be a little

01:00:52.420 bit immodest with, which is how do you do what you do? I don't meet a lot of people that I look at

01:00:59.020 and realize that on every level of their life, they are better than me. You know, usually I meet

01:01:04.640 people and it's like, yeah, they're better at me at those three things, but I can do this one

01:01:07.860 thing better. And I'm sure I could think of something I do better than you, but it's-

01:01:11.560 Many things. You don't have my sense of balance.

01:01:14.360 I can probably shoot a bow and arrow better than you. But when I think about how you balance

01:01:19.680 the devotion you have to your patients, to your family, to your research, to your writing,

01:01:26.220 it humbles me. I don't know. Have you put any thought into that? The fact that you created these

01:01:30.360 tenets around writing is very interesting to me. It says that you weren't just a natural gifted guy

01:01:35.940 that fell out of the sky who figured out how to write. You know, you had to work at your craft.

01:01:40.020 So how do you work at this craft of just excellence in general?

01:01:43.960 The simple rules that I have is I'm a very question and project-driven person.

01:01:48.240 If I set projects, I'll usually fulfill them. Again, it's the same sort of crampon in the

01:01:53.100 mountain rule, which is that in order to write a book, you have to write the first line of the book.

01:01:57.420 And inevitably, that's not going to be the first line that survives. In order to do research,

01:02:01.880 you have to do one experiment. And inevitably, that experiment is going to not work out. You're

01:02:05.620 going to do 10, 15 iterations of it, etc. To me, the fundamental rule that works for me is just

01:02:12.580 to throw something at the world, the first line, the first experiment, the first idea,

01:02:16.820 and then keep at it. I keep at it over and over again. I come back to it. I come back to it over

01:02:22.260 and over again. And I keep asking questions. And then at some point of time, and that's another

01:02:27.920 moment of sensitivity, is to figure out when the work speaks back to you. The experiment starts

01:02:32.900 talking back to you. You have to be really open. Your ears have to be really open. And that's

01:02:37.100 really the skill of a scientist, I think. A scientist, I think, or even a writer, they have

01:02:42.660 really two skills that they mix together. The first one is the creative step, putting out the first

01:02:49.440 line, thinking of the idea. I'm going to write a biography of cancer. I'm going to write the history

01:02:54.600 of genetics, etc. But the second one is to be open enough to realize when the work starts speaking

01:03:00.760 back to you and let that happen. Let the experiment start talking back to you. Be skeptical of it.

01:03:05.360 Have a conversation with it. Those, I think, are the two skills that I sort of bring to my puzzle.

01:03:11.260 The rest of it is just like everyone else, trying to balance eating a meal versus, you know, writing

01:03:16.660 in two more lines, stopping, starting, sleeping, the usual.

01:03:21.800 Well, Sid, I don't expect this will be the last time we sit down and do this.

01:03:25.140 I hope not.

01:03:25.260 There's going to be a lot more stuff to talk about. I can't thank you enough. I consider you,

01:03:29.300 Lou, I mean, some of the other folks who talk about, I mean, real mentors of mine. And it's,

01:03:32.900 it's a, it's a privilege. It's a real privilege to call you a friend and to be able to sit down

01:03:37.420 with you and just get even the tiniest insight into your work. Thank you.

01:03:44.360 You can find all of this information and more at peteratiamd.com forward slash podcast.

01:03:49.620 There you'll find the show notes, readings, and links related to this episode.

01:03:53.420 You can also find my blog and the nerd safari at peteratiamd.com. What's a nerd safari, you ask?

01:03:59.880 Just click on the link at the top of the site to learn more. Maybe the simplest thing to do is to

01:04:04.100 sign up for my subjectively non-lame once a week email, where I'll update you on what I've been up

01:04:08.680 to, the most interesting papers I've read, and all things related to longevity, science, performance,

01:04:14.020 sleep, et cetera. On social, you can find me on Twitter, Instagram, and Facebook,

01:04:18.020 all with the ID peteratiamd, but usually Twitter is the best way to reach me to share your questions

01:04:23.600 and comments. Now for the obligatory disclaimer, this podcast is for general informational purposes

01:04:28.560 only and does not constitute the practice of medicine, nursing, or other professional health

01:04:32.940 care services, including the giving of medical advice. And note, no doctor-patient relationship

01:04:38.300 is formed. The use of this information and the materials linked to the podcast is at the user's

01:04:43.680 own risk. The content of this podcast is not intended to be a substitute for professional

01:04:47.820 medical advice, diagnoses, or treatment. Users should not disregard or delay in obtaining medical

01:04:53.340 advice for any medical condition they have and should seek the assistance of their health care

01:04:57.540 professionals for any such conditions. Lastly, and perhaps most importantly, I take conflicts of

01:05:03.280 interest very seriously. For all of my disclosures, the companies I invest in and or advise,

01:05:08.440 please visit peteratiamd.com forward slash about.

The Peter Attia Drive - December 10, 2018

#32 - Siddhartha Mukherjee, M.D., Ph.D.： new frontiers in cancer therapy, medicine, and the writing process

Episode Stats

Summary

Transcript