The Peter Attia Drive - #346 - Scaling biotech and improving global health： lessons from an extraordinary career in medicine ｜ Susan Desmond-Hellmann, M.D., M.P.H.

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

00:00:16.540 my website, and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:21.520 into something accessible for everyone. Our goal is to provide the best content in health and

00:00:26.720 wellness, and we've established a great team of analysts to make this happen. It is extremely

00:00:31.660 important to me to provide all of this content without relying on paid ads. To do this, our work

00:00:36.960 is made entirely possible by our members, and in return, we offer exclusive member-only content

00:00:42.700 and benefits above and beyond what is available for free. If you want to take your knowledge of

00:00:47.940 this space to the next level, it's our goal to ensure members get back much more than the price

00:00:53.200 of the subscription. If you want to learn more about the benefits of our premium membership,

00:00:58.020 head over to peteratiyahmd.com forward slash subscribe. My guest this week is Dr. Susan

00:01:06.580 Desmond Hellman. Sue is a physician who is board certified in internal medicine and medical oncology.

00:01:12.780 Her impressive career has spanned multiple fields. She has been a leader in the pharmaceutical industry

00:01:17.640 where she helped develop several groundbreaking drugs, worked as the chancellor of the health

00:01:22.200 science campus of a major university system, UCSF, and served as the CEO of the Bill and Melinda

00:01:28.240 Gates Foundation. She also served on numerous boards of both corporations and non-profit organizations.

00:01:33.960 She co-chaired the National Academy of Science Committee that pioneered precision medicine and

00:01:38.500 currently sits on the board of OpenAI. I wanted to have Sue on this podcast to speak about her 1.00

00:01:43.500 extraordinary career spanning medicine, oncology, biotech, and global health leadership, and to really

00:01:49.240 explore her knowledge on how scientific innovation and leadership can drive better healthcare outcomes.

00:01:54.840 In this episode, we discuss her early days in medicine training at UCSF during the start of the

00:02:00.000 AIDS crisis before people even knew what it was and the lessons that she learned on handling uncertainty,

00:02:05.860 balancing public health messaging, and accelerating treatment breakthroughs.

00:02:08.980 The decision that she made to specialize in oncology and how her time treating HIV-related

00:02:13.580 cancers in Uganda reinforced the need for integrating epidemiology patient care and policy to combat

00:02:19.740 global health crises. We spoke about her transition into biotech, helping develop breakthrough cancer

00:02:25.300 drugs like Taxol, Herceptin, and Avastin, and the role of precision medicine in improving outcomes.

00:02:31.360 Sue talks about her leadership roles at UCSF and at the Gates Foundation, driving innovation in healthcare

00:02:35.940 and global health, and the lessons learned from leading health research institutions and global health

00:02:41.740 initiatives, balancing financial constraints with scientific progress, and building culture.

00:02:47.100 We end this discussion with a perspective on the future of medicine, including AI's role in healthcare,

00:02:51.900 such as the opportunities and challenges in leveraging AI for drug development, diagnostics,

00:02:58.000 and expanding access to high-quality care. So without further delay,

00:03:01.440 please enjoy my conversation with Dr. Sue Desmond-Hellman.

00:03:04.820 Sue, thank you so much for making the trip out to Austin. Really, really was excited to meet you

00:03:16.000 last year. Just an honor to spend part of a day with you and then realize that I could somehow twist

00:03:21.120 your arm into coming on the podcast. I'm happy to be here. I look forward to it.

00:03:24.560 You've had just an unbelievable career. You are an absolute giant in many ways. I love to always give

00:03:30.480 people a sense of how someone got to where they got. So if I recall, you grew up in Reno. Is that

00:03:35.580 right? I did. Yes.

00:03:36.720 And you went to high school and college and even medical school all the way through, right?

00:03:40.860 I went to Catholic school for 12 years in Reno. I explain that when people wonder if I was at a

00:03:45.760 casino for my childhood. And then I went to University of Nevada, both undergrad and to medical school.

00:03:52.120 Then that you ended up at UCSF for your residency?

00:03:54.960 You know this with residencies. My dream residency was internal medicine at UCSF,

00:04:00.480 my first pick. And I got my first pick and went to UCSF as an internal medicine resident.

00:04:06.820 And that would have been what year that you landed there?

00:04:09.760 1982.

00:04:11.020 Okay. So remind me where we were in the AIDS epidemic in San Francisco in 82. What was known?

00:04:17.440 If you read MMWR, that's 1981, was the first indication. In 1982, we knew that there was

00:04:26.980 something happening, especially to gay men, but there was a sense it was homosexuals,

00:04:33.260 hemophiliacs, and Haitians. Remember that? 0.98

00:04:36.000 Yeah.

00:04:36.260 Three H's. There was so much mystery still involved that I was and my colleagues were in a study to

00:04:43.960 look at drying our blood to see if we had been infected as a result of treating patients.

00:04:49.320 And what were they presenting with at the time? It's hard for anyone of even my generation. You've

00:04:55.700 never seen a drug-naive patient. All of my experience with HIV, which was a lot in Baltimore

00:05:00.240 many years later, but everyone was on something. So how would these men present to you as a medicine

00:05:07.020 resident?

00:05:08.300 Pneumocystis. Pneumocystis, coronia, pneumonia was the number one diagnosis. So that's what you saw

00:05:14.140 in the hospital that brought patients to attention. Tell folks why that's so unusual.

00:05:19.880 Oh, it was a disease that immunosuppressed patients could get very rarely. Most clinicians

00:05:25.920 had never seen it before. What was also clear is that there were many other infections that were not

00:05:32.620 as obvious or life-threatening as pneumocystis was when we saw it right away. What was interesting

00:05:39.320 from an outpatient perspective was capsic sarcoma. Tell folks what that is.

00:05:43.980 Capsic sarcoma is a really unusual purple-colored tumor, very visible externally. It caused nodules.

00:05:53.120 In patients with HIV infection, it also caused internal organ involvement. And patients would

00:05:59.700 cough up blood or they would vomit blood. But what was really sad and difficult is the combination

00:06:06.060 of cachexia and capsic sarcoma meant that everybody knew you had AIDS. They sort of wore it. And what

00:06:12.380 was interesting for me was that this old-fashioned capsic sarcoma was fundamentally different than what

00:06:19.520 we were seeing. We also saw non-Hodgkin's lymphoma in numbers much smaller than capsic sarcoma. But

00:06:26.540 capsic sarcoma was a very big problem in San Francisco. It was very common in gay men and it was common

00:06:33.540 in the population we saw. And was there ever a sense of fear among the medical staff that we

00:06:39.360 don't know what this is, we don't know how it's transmitted, and therefore we don't know how to

00:06:43.340 protect each other or ourselves or other patients for that matter? Like, it's hard for me to imagine

00:06:48.080 that given how much we take for granted today. I think it's probably a reflection of my own

00:06:54.320 personality and my own wish to be a physician that my memories of those days are much more about sadness,

00:07:02.720 about my patients, and about people my age dying or being pretty clear they were going to die. I mean,

00:07:09.740 a story that brings it to life is many patients started selling their life insurance because they were

00:07:15.180 sure they wouldn't live long enough and they wanted the money now. And then when the antiretroviral

00:07:20.120 therapy came along, they wished they hadn't, which is a good thing to have. But I was just really sad.

00:07:26.120 There were fears about the residents and about contagion. But in San Francisco, there was such a

00:07:32.780 wish to help the patients and such a good spirit about playing a role in helping that we all persevered.

00:07:41.940 But the first patients I took care of in the hospital, I remember very well in 1982,

00:07:46.260 we were a gown, gloved, masked, had a cap on. It was like we were going into an operating room.

00:07:51.860 Got it. For all intents and purposes, you were acting like this was Ebola without knowing.

00:07:55.480 Absolutely. Absolutely.

00:07:56.660 And so you finished your residency in internal medicine. Did you go directly into your fellowship?

00:08:02.660 I did a chief residency at the university hospital. And I think that was the first that I knew I really

00:08:08.580 liked managing. I really liked interacting with people and helping people succeed.

00:08:12.780 So I did that for a year and then went into my oncology fellowship after that year.

00:08:18.740 And why did you pick oncology?

00:08:20.400 Oh, to this day, I love oncology. If you love medicine, and I do, and you love patients, and I do,

00:08:27.860 it's the combination of you get to call on your compassion gene and your nerdy science gene.

00:08:33.920 And when I was in Reno, at the Reno VA, I had an attending, Stephen Hall. And he was the oncologist

00:08:43.300 who was teaching me about medicine, third-year medical student. And I loved everything about how

00:08:49.600 he showed up. I loved his compassion. I loved his intellect. And after that, I had in my mind this bug

00:08:58.140 about, I wanted to be like him. I can really relate to that. When I was in my third year of

00:09:03.440 medical school, I went to the NCI for three months with Steve Rosenberg. And it was the exact same

00:09:09.480 experience. And I remember learning many lessons from Steve. One of them was that cancer diagnosis,

00:09:15.740 and of course, at the NCI, as you know, nobody's showing up with stage one, two, or three cancer.

00:09:20.260 By definition, every patient there is showing up with metastatic cancer. And they've progressed

00:09:24.340 through all standard treatments. So these are people that have six months to live. And maybe

00:09:28.040 10% of them, you actually come up with a durable remission for. But he said, look, cancer will do

00:09:33.340 one of two things to a family. It will take a strong family and bring them much closer together.

00:09:38.360 It will take a fractured family and blow them wide apart. As a doctor, as a nurse, as an anybody

00:09:44.440 in the field of oncology, your ability to kind of be there for that family is as important,

00:09:50.520 potentially more important than it is in any other specialty of medicine.

00:09:53.580 That's really well said. I love that.

00:09:56.280 So tell me a little bit about the state of oncology in the mid-80s when you're embarking

00:10:01.720 on your medical oncology fellowship. Help people understand what the world of cancer looked like

00:10:06.760 roughly 40 years ago.

00:10:08.240 Let me talk about breast cancer. That's a cancer that is such a good example. The common therapy that

00:10:15.440 was used, cytoxan, methotrexate, 5-FU, were very old, decades old. There were no new chemotherapy

00:10:22.480 drugs. It hadn't been in a while. The field was stifled, I would say, in terms of medical

00:10:28.440 oncology. There wasn't a lot going on. I was really interested in cancer epidemiology. That

00:10:35.060 was something to me that asking the question, why did people get cancer and couldn't we do

00:10:39.600 something about it, seemed really important to me. I wanted in the second year of my fellowship

00:10:45.260 to study the relationship between hepatitis B and hepatocellular carcinoma and to understand

00:10:50.960 that better and to think about the viral link with cancer. The mentor I was supposed to work

00:10:57.240 with ended up not coming to San Francisco. So I decided to go to Berkeley and get a master's in

00:11:03.000 public health as a backup strategy. I really scrambled because I didn't want to waste a year.

00:11:08.040 What was the nature of the program? It was a three-year fellowship with a research track on

00:11:12.760 the side because obviously UCSF is such an academic place. It's a very academic place,

00:11:16.780 but you could do two or three years and many people went into the lab. I didn't want to go into the lab.

00:11:23.520 I wanted to do epidemiology. I wanted to learn more about statistics and epidemiology. I thought I

00:11:30.780 wanted to do it because I wanted to be a cancer epidemiologist. And to this day, I still think that

00:11:37.120 is one of the great opportunities to make a big impact, but you have to be funded. So I'm a

00:11:44.280 pragmatist. The good news was that all that learning at Berkeley and at UCSF in epi and biostat,

00:11:52.780 I brought to drug development. Clinical trials have a lot in common with doing epidemiology.

00:11:59.560 You brought up the example of hep B and hepatocellular carcinoma. Was it understood at the time what we

00:12:04.120 now know? It was, yeah. Palmer Beasley, one of the fathers of that relationship, was the guy who

00:12:10.520 was supposed to come. There were preliminary papers and something relatively early, but it was emerging

00:12:16.420 science. Do you recall what the incidence of hep B was and hep C back then? You know, I don't.

00:12:23.880 If you weren't in Asia, it was actually, I think, relatively low, but I believe increasing, 1.00

00:12:28.840 which is partly why the vaccines are so important. Tell me about how you wound up in Uganda.

00:12:35.360 After I got my master's in public health, I became the oncologist at UCSF in the university hospital

00:12:43.560 for the AIDS clinic. This is Moffitt?

00:12:46.180 This is Moffitt. So San Francisco General had a very well-known program run by oncologists for AIDS

00:12:52.820 patients who were in the safety net hospital. But in the university hospital, if you were very sick

00:12:59.400 and you had capsid sarcoma, you saw me. And my husband, because we had just gotten married,

00:13:04.340 we were interns together, he was in the lab in ID doing immunology work. So two of the chiefs of

00:13:12.100 medicine at UCSF were approached by the Rockefeller Foundation, who had started to become worried about

00:13:18.760 heterosexual transmission of HIV. Remember I talked about the Haitians and the hemophiliacs and 1.00

00:13:24.420 homosexuals? 1H wasn't heterosexual. And so there was a lot of disbelief about African HIV. And in fact,

00:13:32.840 some people thought it must be gay sex, but people are too embarrassed to admit it. There were other 0.98

00:13:38.340 theories, but people just did not understand what was going on in Africa. So the Rockefeller said,

00:13:44.180 we'll give you a grant at UCSF. We'll grant you money to study heterosexual transmission of HIV.

00:13:50.680 And this was through an epidemiologic contact tracing lens, not necessarily going into the lab

00:13:55.640 and trying to figure this out. Not going to the lab, but really looking at epi. And particularly,

00:14:00.520 there was a hypothesis that if it was heterosexually transmitted, there was something to do with

00:14:05.380 sexually transmitted diseases. And that there was something about increasing your risk if you had

00:14:10.660 untreated STDs, sexually transmitted diseases. So we were asked to go. UCSF had no global health.

00:14:18.160 To put this into context, we had a flat and two Honda Civics. I still remember this. We gave my dad

00:14:24.640 power of attorney. We sublet our flat and we sold our Hondas and moved to Uganda. I'm laughing in part

00:14:31.320 because I had never been east of Chicago. I mean, this was a pretty dramatic thing to do. And it was

00:14:39.360 only... And I'm sorry, you, your husband, and who else? The two of us. That's the team. That's the

00:14:44.000 dream team. That's the team. That's the team. And Uganda was a place where on the positive side of

00:14:50.080 things, the NCI had set up a collaboration with Uganda Cancer Institute, where they did some really

00:14:55.960 great things in lymphoma and Burkitt's lymphoma, if you remember those stories. And one of the

00:15:02.060 physicians at UCSF had been associated with that, John Ziegler. So there was a connection to the Uganda

00:15:07.380 Cancer Institute. So on the good side of things, there was that. And there also was and is the

00:15:13.260 Entebbe Viral Institute. So there was some infrastructure there. Unfortunately, most of that

00:15:18.540 infrastructure had been ruined by the idiomine regime not long before we went to Uganda. So when we

00:15:25.920 went there, it was pretty lawless. There were roadblocks you had to stop at. It was difficult

00:15:31.680 to live there. It was really difficult. And what about safety? I would say now that I'm used to being

00:15:37.860 in more safe situations and older and wiser, it was probably not that smart the way we lived there,

00:15:44.360 but we weren't reckless. It seemed dangerous when you were in the car to have carjacking or your money

00:15:52.160 go missing or things like that.

00:15:54.000 Was Idi Amin still ruler?

00:15:56.160 Idi Amin was gone, but when we were there, he made that attempt to come back from Saudi Arabia and go

00:16:01.700 back to Uganda, but it was thwarted. So that was good news. So Nick, my husband, reestablished the 1.00

00:16:08.280 Sexually Transmitted Disease Clinic and attended in the internal medicine ward. And I like to say I

00:16:14.520 doubled the population of oncologists in Uganda when I was there. So my colleague, Edward M. Biddy,

00:16:21.640 who's Ugandan, put all his focus on the pediatric unit and I put all my focus on the adult unit,

00:16:28.420 which was so many cases of capsic sarcoma.

00:16:32.160 Give me a sense of what this meant. So we're talking late 80s now.

00:16:35.340 This is 89, 90, and 91.

00:16:37.560 Is AZT out yet?

00:16:38.820 Not yet.

00:16:39.400 Okay.

00:16:39.600 Just on the brink.

00:16:40.620 Okay. So we have nothing. And what is the approximate conversion? So for a patient who

00:16:47.560 develops AIDS, what fraction of those will go on to develop KS?

00:16:52.440 If you were in Uganda at the time, gosh, especially amongst males, but also males and females, 0.93

00:17:00.300 it's so hard to give those numbers. But I would say about a third of patients who

00:17:04.780 sought medical attention probably had KS, some KS.

00:17:08.340 What was the prevalence of HIV AIDS in the population in Uganda?

00:17:13.440 Depend on the population you treated. It was double digits in the country as a whole.

00:17:18.780 If you were 16 years old, if you were a 16-year-old girl and you went to the STD clinic,

00:17:24.740 you had a 50% chance of being HIV positive. 16. And most of those girls was their first

00:17:31.100 and only sexual partner. It was Russian roulette to have sex in Uganda then. 0.97

00:17:36.380 I mean, worst Russian roulette's one in six if you've only got one bullet in the chamber. 0.96

00:17:40.400 You got the bullet in one of the two chambers. Yeah. Yeah. And the best business in town,

00:17:47.080 coffin maker. We would go, we would drive back to where we stayed and you would see if you've

00:17:52.680 ever been in an African village, like they'll prop up the coffins made of wood and you just 1.00

00:17:57.480 see them because that was so astounding. The feeling of being scared and sad in San Francisco

00:18:03.620 in 1982, multiply that by a thousand in 1989. It was terrifying. If we hadn't gotten ARVs,

00:18:13.820 this was killing people. But you know, the same time, the first time we went back to San Francisco

00:18:20.480 from Uganda was six months after we had left. I went back to the capacity sarcoma clinic that I

00:18:27.140 had led and said to the nurse, oh, yeah, ask about your patients. I had so many great guys who I cared

00:18:33.900 for. All my patients were dead. All of them. Six months. The sense of how bad HIV was before

00:18:42.320 antiretrovirals. It's impossible to overstate it. Just impossible. And when we were in Uganda,

00:18:49.840 it was really clear that you could see someone's immune status with a good physical exam if they

00:18:57.100 had capacity sarcoma. I wrote a paper that I think is a good paper if you do global health and you have

00:19:02.820 limited resources. It was a paper that had one observation. If you had capacity sarcoma on your

00:19:09.160 soft palate, on the roof of your mouth, you had HIV. 100% predictive. Capsisarcoma,

00:19:16.280 there's a Mediterranean form and an African form. It happens on your skin. It can cause elephantiasis,

00:19:21.960 but it doesn't go in. The mouth is just a surrogate for your GI tract. Doesn't happen unless 0.94

00:19:27.800 you're immunosuppressed with HIV. These patients weren't necessarily dying from the KS directly.

00:19:35.380 That's a proxy for how weak their immune system was. I assume they were ultimately dying from a

00:19:40.000 pneumonia? Many would die from pneumonia. There was severe cachexia, and then they were prone to

00:19:45.700 pneumonia and other problems. But Kaposi sarcoma in the lungs or the stomach can also cause bleeding,

00:19:54.360 and you can die from that. What did you know at this point in time about HIV? Because the virus had

00:20:00.020 been identified by this point. What was known and what was unknown? We knew most of the clinical

00:20:05.240 syndromes associated with HIV. This was Gallo? Was it Gallo? Yeah, Bob Gallo was one of the...

00:20:11.000 Luke Monnier was... Yep, yep. Okay. You know, they had a fight over who...

00:20:14.360 Who deserved the credit for that. Who deserved the credit. Yep.

00:20:16.240 But yeah, we knew about HIV then, and we knew the biology, and we knew as soon as we got to Uganda

00:20:22.380 and examined patients, that this was heterosexual transmission of HIV. And we knew that untreated

00:20:28.620 STDs were a big reason, and that was a very important thing.

00:20:33.600 Going back to these 16-year-old girls, is the reason that the sexual... heterosexual transmission 0.98

00:20:39.340 was so high because the viral loads were through the roof? Because today, if a male with HIV had

00:20:45.540 unprotected sex with a female, it would not be that high, would it?

00:20:48.460 No, it wouldn't be that high. No. So one of the really important aspects of STDs is high frequency 0.78

00:20:56.440 of herpes and chancroid, really open lesions that are very, very, if not treated...

00:21:03.900 So it's the one-two punch. Yeah.

00:21:04.920 Super high viral load.

00:21:06.360 High viral load and transmissible.

00:21:07.960 And opening. Yes, yes. So we knew all of that. Now, we also knew that some of these were treatable,

00:21:16.060 that both medication, also Museveni, the still leader of Uganda, had this very funny campaign

00:21:24.760 called Zero Grazing. So they raised a lot of cows, and this is very important in Uganda,

00:21:30.980 is having a herd of cows. It means you're an important man, you know. Museveni always wears

00:21:35.640 this hat like he's raising cows. So Zero Grazing, the farmers and many people knew what that meant.

00:21:40.820 One wife, one partner, no grazing. And so there was a pretty good public campaign. We did a lot

00:21:47.840 of condom distribution.

00:21:49.860 And so the government was receptive to this.

00:21:52.000 Yeah.

00:21:52.220 They understood the science.

00:21:54.540 They understood the epidemiology. And they were completely on board with the campaign.

00:21:58.680 They were very on board. They also knew that this was going to be a geopolitical problem for them

00:22:04.440 if people were dying in the prime of their lives at the rates they were. They got that. This was

00:22:09.940 really clear to them.

00:22:11.620 What other countries in Africa were afflicted to this extent?

00:22:14.380 In East Africa, there was quite a bit. There was a lot of HIV in Kenya, and there were programs

00:22:19.420 like the one that we had in Kenya, Tanzania. There were others where it was more unknown,

00:22:24.640 I think not talked about. I mean, the program I know about most today is the program my husband's

00:22:32.540 been working with for 15 years, which is Elizabeth Glaser Pediatric AIDS Foundation. They work now,

00:22:38.740 I think, in 12 countries in sub-Saharan Africa. And many of the southernmost countries are heavily

00:22:45.040 affected by HIV still.

00:22:46.940 Can you estimate in a year how many people died from AIDS in Uganda when you were there?

00:22:51.340 Oh, no, I can't estimate it.

00:22:53.180 I guess the point is it's a staggering number, and yet there were so few of you that were on

00:22:58.160 the front lines.

00:22:59.720 If there's 16 million people, it wouldn't have surprised me if there were a million people

00:23:03.300 who died. I mean, it's that kind of numbers. I'm probably exaggerating, but not by much.

00:23:08.940 And I think the sense of feeling overwhelmed is just really important. What I realized I was doing,

00:23:16.220 I don't know if you've interacted with people in the military much, but if they were on the

00:23:20.620 battlefield, they triage. I triaged. I triaged in San Francisco. If you didn't need chemotherapy,

00:23:26.560 but you had capsic sarcoma, I didn't see you.

00:23:28.940 What was the chemo?

00:23:30.060 The simple one was vincristine. Vincristine is actually reasonably good against KS. I used it

00:23:36.820 in Uganda a lot. It does cause some neuropathy, but if you're careful about how much. And then

00:23:41.900 gliomycin. Again, you have to be careful because of the pulmonary toxicity. Good old-fashioned

00:23:46.280 vincristine and gliomycin. And then Texel. Texel was approved for capsic sarcoma after I left

00:23:52.580 Uganda. It wasn't a drug before then. I would see the patient, and I would literally ask them and

00:23:58.160 their family, can you walk? If you can walk-

00:24:00.560 If yes, you're too healthy for me.

00:24:01.980 You're too healthy. It will delay. There was triage because I only had on the shelf

00:24:05.500 a certain amount of chemotherapy.

00:24:08.380 How did you manage the personal toll of the grief and the death of seeing this? I mean,

00:24:13.780 look, I think every doctor to some extent goes through this, where you try to sort of compartmentalize

00:24:18.760 what you're seeing. But the truth of the matter is virtually no doctor can really comprehend what

00:24:24.060 you're describing there. How did you process that?

00:24:27.560 I have this philosophy, which I don't recommend it for others. It's just my philosophy. I love people.

00:24:34.260 I love interacting with people. I love getting to know the patients who I care for. And it makes

00:24:41.740 me happy to think I'm helping. Helping might be helping them get better. Helping might be helping

00:24:47.660 with their pain. Or they can talk about dying with me because it doesn't make me scared. So I get a lot

00:24:54.260 of joy in trying to contribute, even if I feel overwhelmed and if I step back and think, how can we cope

00:25:00.660 with this? My coping is... Is leaning in.

00:25:03.600 Yeah.

00:25:04.380 Does your husband share that? Was there a yin and a yang to the relationship where you supported each

00:25:09.400 other in a way that was helpful in that? I do understand what you're saying, and I appreciate

00:25:13.640 that there is a joy that comes from helping people. But I can also at least personally say that

00:25:18.380 there are moments when it breaks down and you feel so overwhelmed by sadness.

00:25:22.560 Well, first of all, my husband is more introverted and probably gets more sad. But we are also

00:25:29.540 a good team because we're there for each other. And I think it's a special thing done in small

00:25:36.480 amounts, not too much to be able to come home and say, boy, that was tough. Here's what I

00:25:41.260 dealt with today. Or I need to tell this story. Or I want to talk about this. The other thing

00:25:46.440 we did, which is I think so important, is I do drive a lot of joy in trying to help

00:25:52.560 but I'm not a martyr. I don't believe in it. Okay, you worked hard, I worked harder. You

00:25:56.780 suffered, I suffered more. I hate that. So we went to Greece. We still laugh about going

00:26:03.380 to Greece and eating our way through Greece for a week. When we were in Uganda, we had

00:26:08.880 a couple of other good trips. We went on a hilarious safari to a place that was Moya Lodge that had

00:26:15.800 been closed to all tourists, had just reopened. And it was so great. We saw hippos and elephants

00:26:21.920 and we realized we were the only, what you call in Uganda, Amazungu, which is a white person

00:26:27.420 there. So it was a grand adventure. So we had some grand adventures and played tennis, enjoyed

00:26:33.320 friends. We did as much to keep our spirits up as one can.

00:26:39.960 And so you came back to the U.S. after about three years. And did you go back to UCSF?

00:26:44.500 Well, we wanted to go back to UCSF, but we had not kept our academic careers going as much as we

00:26:51.720 should have. We didn't publish enough and they didn't have a global health program or money for

00:26:56.960 us. Taking care of a million people with HIV wasn't enough to justify coming back to UCSF.

00:27:02.100 It actually wasn't. So we said, well, gee, when the chief of medicine outlined for us the plan for

00:27:08.080 us to stay, a large part of it was taking care of patients to pay our way. So we said, boy,

00:27:13.480 taking care of patients, we know what that looks like. So we went into private practice.

00:27:17.220 In San Francisco?

00:27:18.000 No, we moved back to Kentucky where Nick is from.

00:27:20.600 Okay.

00:27:21.080 So we moved back to Kentucky and I was in a two-person oncology practice

00:27:25.300 with a former classmate of Nick's in Lexington.

00:27:28.680 And you were doing at this point oncology unrelated to, not necessarily focusing on HIV and AIDS

00:27:35.780 related cancer, breast cancer.

00:27:37.640 I was doing good old-fashioned American oncology. I didn't take my oncology boards when we went to

00:27:46.160 Uganda because I was in Uganda. So I still sort of laugh about taking the DeVita oncology book

00:27:53.480 with a yellow Sharpie. I reread the big oncology book twice.

00:27:59.160 Is this the DeVita Hellman Rosenberg book?

00:28:01.320 Yes.

00:28:01.560 Yeah, yeah, yeah.

00:28:02.080 Of course.

00:28:02.400 Yeah. It's brown these days, I think. I reread it twice, took my boards and did fine. So I was

00:28:07.720 ready.

00:28:09.240 This is unbelievable. So you're sitting in Kentucky practicing garden variety oncology.

00:28:14.660 Talk to me about what that's like. I mean, that's completely orthogonal to what you've

00:28:18.640 been doing for the past couple of years.

00:28:20.240 It was so, so, so, so different. And Nick was in a practice where he was more like a hospitalist.

00:28:25.420 Somebody would get a fever in the ICU and they'd call that ID group. And my practice was a 0.77

00:28:29.380 two-person practice. It was very classy. I saw a lot of lung cancer. It was Kentucky. 0.72

00:28:33.620 So there's a lot of smoking, a lot of people from Appalachian. I actually like taking care 1.00

00:28:39.320 of patients. So that part I liked, but I really missed intellectual research, collegial stuff

00:28:45.960 that I was used to at UCSF because we had been there nine years by that time because we

00:28:50.640 were still UCSF faculty when we were in Uganda. Nick was called about Bristol-Myers Squibb search

00:28:58.060 for an expert on HIV because they were trying to follow AZT with the next antiretroviral.

00:29:06.440 I think it was DDI and D4T were both in development then. And so they recruited Nick to come and work

00:29:13.280 at Bristol-Myers Squibb out of private practice. And Nick said, I won't come unless you have a job

00:29:19.500 for my wife. And they said, no, we have a nepotism clause. We don't allow couples to work at Bristol-Myers 1.00

00:29:25.280 Squibb. So he said, fine, I won't come. He's a good husband. This is one of our favorite stories

00:29:30.660 because it's a true story. So they called him back and they said, we really, really want you to come

00:29:35.120 because we want this program to do well. And could your wife be a consultant? Would she agree to be a 0.76

00:29:40.660 consultant and not a full-time employee? And he said, yeah, that'll work. So we moved to Connecticut.

00:29:46.220 He had a job and I was the trailing spouse. And I can just see what this looked like. I'm making

00:29:52.460 this up now. Oh God, we've got an LMD. You know what an LMD is. We got this lady from Lexington, 1.00

00:30:00.240 Kentucky. She's in private practice, oncology, and we're stuck with her. Let's have her do drug 1.00

00:30:06.440 safety on Taxol. We have this new drug and it's really busy. It looks like it might work. And so

00:30:11.500 we'll put her on drug safety. She can't hurt anything doing that. But did they not understand

00:30:17.280 what you had spent the last couple of years doing prior to being in Kentucky? Did they not know what

00:30:22.980 you had done in Uganda? I don't think that registered because there were so many people there 1.00

00:30:29.340 who were very traditionally trained at NCI or at Yale or wherever they were. And they were

00:30:36.760 traditionally trained in oncology. My experience in Kampala in Uganda didn't make an impact. But

00:30:43.580 here's what was funny. Nick and I didn't have a statistician. As I told you, we just, the two of

00:30:48.560 us went. So we brought this little compact computer and all the SAS manuals. We didn't have a TV. We

00:30:56.980 didn't have newspapers. We didn't have anything. So we taught ourselves how to do SAS programming.

00:31:01.700 When I got to Bristol-Myers Squibb, one of the really interesting things about Taxol is it causes

00:31:07.220 severe neutropenia, but it's short. It's like this short, severe neutropenia. And so I wanted to study

00:31:14.720 that because I thought it was really important in why people weren't really getting infections.

00:31:18.840 I always have to remind myself, tell people what Taxol is, how it works. Just give them a quick,

00:31:23.040 what is neutropenia? Why would we care?

00:31:24.400 When I talked before about how few new chemo drugs there were, Taxol was one of the first

00:31:30.640 new chemotherapy drugs. So Taxol is a product of the yew tree and it's a microtubule poison. It is,

00:31:39.000 if you think about it coming from the yew tree and you think about sap, think about trying to

00:31:43.300 dissolve sap in water and give that to a patient. That's plenty hard. The dissolving fluid that's

00:31:51.300 given with Taxol. And the reason we would give somebody with cancer a microtubule inhibitor

00:31:56.520 is because that prevents cells that are dividing.

00:31:59.880 You can't divide.

00:32:00.760 They can't, they need these microtubules when they create new cells and we want to block that.

00:32:04.500 And we want to block that. So this was not just a good way to block cellular division,

00:32:09.360 which is so important in cancer therapy. It's really important because it's very different

00:32:13.320 than some of the old chemotherapy drugs. And if you're resistant to those old drugs,

00:32:17.720 here you have a brand new mechanism of action. So that's a terrific thing, but it's not easy to

00:32:23.680 dissolve it. So the dissolving agents are like soap. They dissolve the Taxol. And when the National

00:32:31.520 Cancer Institute tried to use it, some patients got severe allergic reactions from that and they got

00:32:37.660 scared and put it on the shelf. So Bristol-Myers Squibb went to the National Cancer Institute and said,

00:32:42.240 you know, that drug might really be active. We're willing to carefully go back in the clinic

00:32:47.300 and test it and give people agents to counterbalance the allergic reactions and see if we can get

00:32:54.520 away with it. So they did that and they got an approval in ovarian cancer, a brand new agent.

00:33:01.300 Now, Taxol was really exciting because first ovarian and then breast cancer were these indications

00:33:08.020 where we had not had new drugs or really any drugs that were active in the case of ovarian for a long

00:33:13.080 time. And because I was a safety person, I was really trying to understand and put into context

00:33:18.800 all these safety issues so it was possible to safely treat patients with these drugs.

00:33:24.380 It had already been approved.

00:33:25.460 It had been approved for ovarian cancer when I showed up.

00:33:27.980 So now you're doing post-marketing surveillance on safety.

00:33:30.300 We're doing post-marketing surveillance on ovarian and putting together a U.S. submission

00:33:34.920 and a European submission for breast cancer. So I started talking to the statisticians there

00:33:39.740 about how I wanted them to program to get the data we needed for the safety label. And I'll never

00:33:45.140 forget the guy looking at me and saying, do you know how to do this? I said, well, I had to learn

00:33:51.440 in Uganda because I didn't have somebody like you, you know? So it was sort of funny that I was very happy

00:33:58.520 to prove myself. It didn't bug me. It made me more feisty. Like, I'll show you, I'm not underdosed

00:34:05.140 in the kinds of things you need to do in this place. And by the way, I loved every minute of

00:34:10.800 being at Bristol-Myers Squibb. They were pros at cancer drug development. They were pros at monitoring

00:34:17.260 safety. And I thought it was so much fun because you got to make drugs.

00:34:24.100 What was the pharma landscape like in the early 90s? So you had Bristol-Myers, you had Pfizer,

00:34:28.080 you had Merck. You had Merck, you had Novartis. I think Novartis was a combo of a couple.

00:34:33.980 It was smaller, much smaller. And cancer was Bristol-Myers Squibb. I mean, they had made

00:34:39.600 cisplatinum, carboplatinum. They had a lot of those drugs. And people who had made those drugs were

00:34:44.960 still there. And I was really happy to learn from them. I felt very lucky to get to be around these

00:34:51.600 folks who knew about Taxol. So we got Taxol approved in the U.S. and in Europe for breast

00:34:57.960 cancer. It became Bristol-Myers Squibb's number one drug. And I became the project team leader for

00:35:03.540 Taxol.

00:35:04.380 How long did it take them to thank your husband for forcing them to bring you along?

00:35:10.220 Too long. He really enjoys that story because he, like all good family stories,

00:35:17.060 it gets embellished over the years. And he tells this story like, you should actually want her.

00:35:21.760 You don't know this, but he's a good husband.

00:35:25.240 When you pause at where we are in this story, to think of everything that would come from this

00:35:29.600 moment forward, and to realize there's a scenario under which nobody knows everything that's about

00:35:35.140 to happen. And you're an oncologist in Kentucky right now.

00:35:37.640 That's right. That's right. I'd be better at tennis.

00:35:40.600 Yeah.

00:35:41.040 I had more free time. But yeah, no, I think that's the thing that I love to mentor. I think

00:35:48.560 it's really underrated to listen to students and hear what's on their minds. And I remind students

00:35:54.960 about the role of serendipity. And I think I'm a poster child for the role of serendipity.

00:36:00.880 So you left Bristol-Myers Squibb in 95?

00:36:04.940 95.

00:36:05.540 And you went to Genentech. Tell me where Genentech was in its life cycle then. So Genentech

00:36:10.940 had been around for a while. I mean, when did Genentech get founded? In the mid-70s?

00:36:15.140 1976.

00:36:16.040 Okay. Give folks a little bit of a history of Genentech. Genentech's a storied company,

00:36:20.600 but also a different company in that it was founded on a new technology.

00:36:24.780 Genentech's a really interesting company because it claims to be the first biotech company. There's

00:36:29.900 some Cetus back and forth about that, but it was based on genetic technology. That's where the

00:36:36.740 Genentech name came from. And what Herb and Bob, the co-founders of Genentech, wanted to do is kind

00:36:44.080 of do a proof of concept that you could use genetic technology and make medicines, make big medicines,

00:36:50.460 proteins, antibodies, medicines that would almost certainly have to be injected rather than

00:36:55.200 swallowed because they're large and they're proteins. So you break them down if you swallow

00:37:00.240 them. But their initial goals were focused on insulin, which they out-licensed to Lilly and

00:37:07.640 Pfizer, and growth hormone, human growth hormone. So before Genentech, when you were a parent and

00:37:14.940 your child was short, you needed to give that child growth hormone that came from cadavers. And that 0.99

00:37:21.540 had a risk of this slow virus disease, and that was not a good trade-off for parents. So the concept

00:37:29.620 of having recombinant, of having human-like growth hormone was a really wonderful thing.

00:37:35.020 So Genentech's first drug was human growth hormone, and it was a tour de force. It was really amazing

00:37:42.700 that in the late 70s, they were able to do this fermentation and purification because they had

00:37:49.420 to prove to FDA it was pure human growth hormone with no contaminants. And it became famous for that,

00:37:56.240 and people were excited and thought this was cool.

00:37:59.240 Tell people briefly how this worked. What was recombinant DNA technology? What were they putting

00:38:04.100 the gene into? How did they get the gene to make the protein? We take this all for granted today

00:38:10.340 because we have- It's tricky, yeah.

00:38:11.920 Yeah, but it's so incredible.

00:38:14.540 So what you see if you go to Genentech or a company like Genentech is you see these tanks,

00:38:19.700 and the tanks are like a cell ICU, like an ICU for a cell. So the cell, what you're doing is you're

00:38:28.400 teaching the cell to make at very high amounts growth hormone, way higher than your cells or

00:38:35.280 my cells would. And then you're teaching the cell through this genetic engineering to secrete it

00:38:40.780 into a medium, into this soup that is really a lot of growth hormone. And then you're taking away

00:38:47.900 the cells after they secrete it, you're purifying that growth hormone, and you put it in little

00:38:53.520 vials. And that's the process of biotechnology. And you can trick a cell into making almost anything

00:39:00.460 you want, not completely, but almost anything you want, and make it very much like human,

00:39:06.020 which is neat. So you don't have to go to a human to donate you growth hormone because it would be

00:39:11.320 too small.

00:39:11.920 Or in the case of insulin, I mean, they were using insulin from pigs and-

00:39:14.900 porcine insulin that made allergies and expensive. So human insulin really changed how you thought

00:39:21.280 about treating people who have diabetes. So Genentech made growth hormone and Genentech

00:39:27.960 sold growth hormone and set up something I actually think is a really neat thing that Genentech did,

00:39:33.480 which people said, how do you know that by giving kids extra growth hormone, it won't cause leukemia

00:39:38.980 or a fourth arm to sprout out or something, you know, weird things to happen. And Genentech said,

00:39:44.160 well, we'll follow every child. So they set up a patient registry, one of the first patient

00:39:48.980 registries ever, and followed every child until they reached their final adult height. And the

00:39:55.300 physicians and their staff entered this into a computer. And so this is an amazing amount of

00:40:02.020 information. So if the FDA ever asked us, do you have this with growth hormone? Did you have that

00:40:06.820 with growth hormone? We had not an example, not, you wouldn't even do statistics on it. Every child

00:40:12.600 ever treated with Genentech's growth hormone. Do you have a sense of how many kids that was?

00:40:16.740 Oh, hundreds of kids, thousands of kids by now. Yeah. Yeah. So Genentech got really good at that.

00:40:22.640 And in fact, when I went to Genentech in 1995, the chief medical officer of Genentech was a

00:40:27.800 pediatric endocrinologist, an expert on short stature and growth hormone. But it's a pretty small

00:40:33.240 market. This is uncommon. By this point, it was being used rampantly in sports.

00:40:38.420 Yeah. And the FDA was not happy about that and pushed really hard on Genentech to control that

00:40:44.820 use. Was it being used by this point also pretty heavily in HIV, right?

00:40:48.480 People were using it in HIV. They were using it in sports. Anything where you wanted to

00:40:52.100 have more muscle mass. That's exactly right. Genentech had done some studies to look at whether

00:40:57.940 that was a good idea. And none of the studies came out successful.

00:41:02.020 Meaning there was no benefit to an HIV patient being on growth hormone?

00:41:06.620 The benefits did not outweigh risks of having increased blood sugar and some other things

00:41:12.680 that would happen. So one of the aspects of Genentech that happened in the early years before

00:41:17.260 I was there is they learned how to make enzymes. Same genetic technology, telling the cell, make

00:41:22.940 these enzymes. And some of the enzymes actually got out-licensed to make commercial enzymes like

00:41:29.500 that you use when you wash clothes and things like that. So that wasn't core to Genentech.

00:41:34.940 But they had an enzyme activase, a TPA, tissue plasminogen activator, that could break down blood

00:41:41.820 clots.

00:41:42.660 Did they go after that knowing what they were doing? Or was this a bit of a fishing expedition

00:41:46.680 where they realized in the process of trying to do many things that, oh my God, we can actually

00:41:52.680 make TPA, which you're going to explain in a minute why that changed the game of cardiovascular

00:41:56.560 medicine?

00:41:57.780 It was intentional. They had a really great, there's a clinician researcher, Dave Stump,

00:42:03.680 who is a clotting expert. He's hemonc on the heme side. He was there and really pushed them to do this.

00:42:10.720 And the concept was that if you could break down the blood clot, you could cure the heart attack.

00:42:16.820 You could save lives. And the interesting thing, if you are interested in doing trials, is they

00:42:22.820 started the concept of a large, simple trial. This was early on and people in cardiovascular,

00:42:28.640 Gene Brunwald and his followers had started these large, simple trials. So Genentech kind of bet the

00:42:34.980 farm on this TPA. And the farm was that they could change the outcomes in 30 days. There'd be more

00:42:42.240 people alive than dead if they were treated with Activase. And two of the people involved in the

00:42:47.540 studies, you probably know their names, were Rob Califf and Eric Topol.

00:42:52.700 Was Eric at Scripps at that time or where was he?

00:42:54.720 I think he was at the Cleveland Clinic.

00:42:56.240 Cleveland Clinic, yeah.

00:42:56.720 And they ran a group called the Timmy Group and they did all these studies named Timmy.

00:43:01.120 And so they did this big trial and it worked. If you treated with Activase, you could break down

00:43:07.000 the blood clots. So Genentech started this franchise in cardiovascular and again, did this

00:43:12.660 really interesting patient registry to look at 30-day outcomes for post-marketing, but stints came

00:43:18.600 along. And so the franchise of Genentech and people who were treated with TPA really went down.

00:43:25.540 And the stroke indication was tricky.

00:43:27.600 You had to ensure it wasn't hemorrhagic or you could make things so much worse.

00:43:32.020 And so the stroke indication on paper was really cool, but pragmatically was really tough for

00:43:37.700 hospitals to execute. Genentech also made another enzyme-like molecule, DNAs,

00:43:44.980 Pulmozyme for cystic fibrosis. And that was approved very tiny. It decreases how thick your

00:43:51.720 secretions are. But with Vertex's CF drugs, it's also been scooped. So when I came in 95,

00:43:59.500 Genentech was really struggling. They had those three drugs. They had growth hormone, TPA, and

00:44:05.440 Pulmozyme.

00:44:06.460 Why did they out-license insulin?

00:44:08.360 I think they needed the money.

00:44:09.620 Got it.

00:44:09.960 I don't think they have the scope to even make it.

00:44:11.700 And why did you decide to leave? Bristol-Myers Squibb is just crushing it. You finally earned the

00:44:17.160 respect you deserve. You've got this struggling company, Genentech. Was it the opportunity?

00:44:24.120 Oh, yeah. Yeah, for sure it was the opportunity. I will tell you, if you were me in 1995,

00:44:30.240 sitting down with Art Levinson, and he was the head of research then, and he was talking about

00:44:35.180 the future and oncology, what the plans were, you'd have gone too. You'd got too. For sure it was

00:44:41.300 an opportunity. We were doing well. I will say that further down on the list of pros and

00:44:47.100 cons was West Coast's home. I mean, Connecticut was snowy and cold. I didn't come over on the

00:44:53.180 Mayflower, it turns out. I mean, I loved people at Bristol-Myers Squibb. I loved the job.

00:44:58.260 But heading back to San Francisco.

00:44:59.440 In San Francisco. Yeah. That was a big deal. But I believed Art when he said,

00:45:04.000 we're going to be a cancer company.

00:45:05.440 So what was the first thing you worked on?

00:45:09.960 Thrombopoetin. They hired me to work on thrombopoetin. It was going to be the third leg of

00:45:14.940 the stool. EPO, so make your red cells go up. Neupogen for your white cells, and TPO for your

00:45:22.360 platelets. And it was a big race. Amgen was in the race.

00:45:26.260 Who developed EPO?

00:45:27.400 Amgen.

00:45:28.020 Amgen. Okay.

00:45:29.060 Yeah. EPO and Neupogen.

00:45:30.240 Got it.

00:45:30.840 Amgen and Genentech had always been kind of rivals. And when they cloned thrombopoetin at Genentech,

00:45:38.280 I read the paper. And then they called me, did I want to come work on it? It was that kind of thing.

00:45:42.440 When you clone it, if you publish it, that doesn't give you the right, it's a race for

00:45:46.920 everybody. It's a race.

00:45:48.340 So let me ask a silly question. Why do you publish the results of the cloning before you've gone and

00:45:54.580 made the recombinant protein yourself? So you patent it, then you publish,

00:45:59.180 then you make the recombinant. Okay. So once it's published after the patent,

00:46:03.260 you get to make it. Mm-hmm. Genentech, one of their great assets started by her boyer is

00:46:09.160 they publish. They don't stop the scientists from publishing. They get the lawyers and they're

00:46:13.420 quick and they make sure that they protect the IP of the company, but they want people to publish.

00:46:17.720 Very academic. So thrombopoetins, EPO and Neupogen are as if you design them to make recombinant

00:46:25.600 forms and give them for cancer patients or other patients who need them. Thrombopoetin, not so much.

00:46:30.800 To make it simple, if you said, okay, your platelets are going way down from your chemotherapy

00:46:37.740 and I'm going to give you thrombopoetin to make them go up, they come back up really late

00:46:44.600 and they go too high. So I'm making you at risk for a blood clot by giving you a million platelets,

00:46:50.980 but later than you need to and you're recovering on your own already.

00:46:54.640 Was that known only once you started developing and you understood the kinetics of it?

00:46:58.840 Once we looked at how it worked in patients, we knew better than we had before. The kinetics

00:47:05.360 of really recovering from not all chemotherapy, as you know, causes your platelets to go low.

00:47:10.420 Yeah. So you can't give it prophylactically because you don't know who's going to get

00:47:13.300 thrombocytopenia.

00:47:14.200 Right. Right. And this tricky thing about going too high, if you're wrong, it's a problem.

00:47:19.280 What do you do? You plasmapherese the patients if you've overshot or not plasmapherese,

00:47:23.140 platelet pharese?

00:47:23.640 You could platelet pharese them. I mean, there are remedies, but you don't want to do that.

00:47:27.320 Anyway, so thrombopoietin proved to be very, very difficult drug. And I learned a lot about

00:47:34.680 the cancer equivalent or the product development equivalent of tulip mania.

00:47:40.720 When everybody's so excited, you get excited too. And it's like, oh, I did learn a lot. I've often

00:47:46.920 reflected on what might happen that I'm not thinking about now. But what also happened is

00:47:52.280 the labs at Genentech had been working on Herceptin, on trastuzumab, for a while. Art became the CEO.

00:48:02.840 Art Levinson became the CEO in 95. And he wanted to push on having trastuzumab, Herceptin,

00:48:09.960 get into the clinic.

00:48:11.520 Tell folks how that drug worked, what it was for.

00:48:13.560 Trastuzumab, or I'll call it Herceptin because it's less of a mouthful, is a antibody. Like you

00:48:19.860 and I have antibodies that fight disease in our bodies. And it's an antibody that targets this

00:48:27.020 protein called HER2. And HER2 matters because about one in four women with breast cancer have too much 0.50

00:48:35.000 of it. And when you have too much of it, if you've got too much HER2 from the time you're diagnosed,

00:48:41.780 your median average survival is three years. If you don't have too much of it, it's seven years.

00:48:48.320 So you know what matters. So the concept with this antibody is turn that off. Whatever bad thing that

00:48:54.300 drives it down to three years, turn that off and go back to seven years. Pretty simple concept.

00:48:58.780 Why do you mechanistically, do you think that the overexpression of HER2 was impeding immune

00:49:04.000 clearance? What was the thesis at the time for why overexpression of HER2 was cutting life expectancy

00:49:10.040 down?

00:49:10.460 The thesis was that it was telling the cell to grow, that it was giving a growth signal to the

00:49:17.300 nucleus to say, grow more. And if you could shut that off, you'd grow less. Now, later we armed

00:49:24.780 HER2 and we put a payload on it. So then you could say, both change the grow more signal and you've

00:49:31.620 got a little bomb on there and you kill the cell. So you'd get a twofer. In fact, another company has 0.60

00:49:37.480 one, AstraZeneca, that's so powerful with a bystander effect, you don't even have to have

00:49:42.800 overexpression. So that antibody, I'm talking about it now because anybody in breast cancer knows about

00:49:48.920 her too.

00:49:49.280 Of course, yeah.

00:49:49.860 You see it on TV and direct-to-consumer ads.

00:49:52.400 Yeah, but I think what I really enjoy about this type of discussion though, Sue, is one,

00:49:56.840 it's the story of your career, but it's also the story of oncology.

00:49:59.600 It is, yeah.

00:50:00.060 It's the story of modern oncology. So you're one of the few people whose careers

00:50:03.800 takes us through the walk of modern oncology.

00:50:07.800 I mention that because it seems impossible. There was a lot of people at Genentech who were

00:50:14.820 negative about Herceptin.

00:50:16.540 Tell me why.

00:50:16.680 I do not think we should invest. The dogma was that antibodies were all hype. They'd been

00:50:23.680 over-promised as smart bombs, smart missiles, Time magazine, all of this, but that they had

00:50:31.180 flopped.

00:50:32.080 What was the biggest failure at that point commercially?

00:50:34.680 I don't think the things had even been commercialized. I don't think they had gotten out of the clinic,

00:50:38.560 that people just weren't seeing benefit. I have a very good friend who's an oncologist and

00:50:45.400 he said, you just can't treat a solid tumor, a solid tumor versus leukemia or lymphoma with an

00:50:51.900 antibody. You need something more powerful. And remember what was happening at the same time is,

00:50:56.960 if you want to talk about the history of oncology, the amazing thing is being at ASCO,

00:51:03.900 the American Society of Clinical Oncology, and two different rooms. One room, we hear that Herceptin

00:51:10.720 is going to change forever how we think about antibodies in breast cancer. Way better than

00:51:15.700 we thought. Improved survival. The other room, doing bone marrow transplant for breast cancer

00:51:22.400 and having the South African group who published a paper saying it worked, retract the paper and go

00:51:28.440 through and talk about how much of it was fraudulent. Fraud, fraud, fraud. So at the same time,

00:51:35.040 this nearly toxic, nearly lethal bone marrow transplant for breast cancer was debunked at the

00:51:42.840 same time as we said, what we call now a naked antibody. No payload, no chemo. You give Herceptin,

00:51:49.920 you're going to help that patient with breast cancer. Just an antibody. Welcome to modern oncology.

00:51:54.880 It could not have been more clear. I almost forgot the BMT stuff. It's so archaic.

00:52:01.580 And it was a distraction because people felt like you just needed to hit the cancer hard.

00:52:09.940 You just needed to hit the cancer smart. Hard wasn't the point.

00:52:14.540 Yeah. I mean, we've got to be getting close to Avastin now too, right?

00:52:17.600 Yes. But don't forget Rituxan. So when I said people didn't believe in you could treat a solid

00:52:25.340 tumor, they thought you could treat lymphoma because we did. Antibodies were so disliked.

00:52:32.640 People did not believe in antibodies that in 95, 96, IDEC was going to run out of money. So IDEC had

00:52:41.140 made an antibody to CD20, a very important marker on all lymphomas. And they were going to run out of

00:52:49.420 money. So some of our business development folks talked to them about Genentech doing a deal with

00:52:56.380 them on Rituxan. It is impossible to overstate how important Rituxan is in lymphoma. I often think

00:53:04.880 when I'm in product development of patients I've cared for, I had this great 83-year-old pharmacist

00:53:10.740 when I was in practice. And he had a lymphoma that was low grade, a little tired. He was fine.

00:53:18.980 And so we did watch and wait, my not favorite strategy of oncology. Let's watch and wait as

00:53:25.440 you dwindle. He'd be a perfect candidate for Rituxan. Four doses. You can repeat it. In fact,

00:53:32.720 it works so well. Here's when I changed my mind on antibodies. Somebody runs in my office and said,

00:53:39.820 oh, we have a case of tumor lysis syndrome. So tumor lysis syndrome being somebody got Rituxan,

00:53:46.040 they had a lot of lymphoma, and the cells are breaking down so fast their kidneys can't keep

00:53:50.920 up and they have to be dialyzed. Oh, that's only an antibody. No chemo, no payload, nothing. That's

00:53:58.260 when you know you've got a good drug. How many cells in the body, how many types of cells in the

00:54:02.160 body express CD20? It's mainly a B cell. But it's not as specific as CD19, is it?

00:54:08.260 I think 19 and 20 are both B cells. I'd have to look at it to see. Yeah, I know CD19 is on the

00:54:13.520 B cell, but I didn't know the, okay. Yeah. We were talking about this earlier. This is chimeric. 0.88

00:54:17.540 It's chimeric, yeah. Yeah. So tell folks what that implies,

00:54:19.940 because that's another wrinkle in the story. Well, that's the other thing that I think is,

00:54:23.600 there's so many dogmas that we believe until data proves otherwise. So one of the warts of

00:54:29.480 rituxan was thought that it was a chimera. It had too many mouse parts to human parts and that we

00:54:34.700 would cause human antichimeric antibody HACA. And FDA was very concerned about this. So we measured

00:54:41.220 and measured and measured. And it turned out probably because the patients had lymphoma that

00:54:46.780 they didn't get HACA. Very tiny numbers and they not clinically relevant. And you can treat them a lot.

00:54:52.200 You can treat patients over and over again. Herceptin is 93% human. So not a chimera,

00:54:58.820 but not fully humanized. And none of the patients treated with antibodies that I've seen,

00:55:04.040 not with Genentech or IDEC antibodies have really had problems. They've had other problems based on

00:55:09.100 target related problems, not based on the antibody.

00:55:12.160 And the CD20 antibody was just also a straight naked antibody?

00:55:15.600 Straight naked antibody.

00:55:16.780 And targeted for an immune destruction?

00:55:18.860 It's targeted to destroy the CD20 positive cells. But it turns out that if you have lymphoma,

00:55:24.820 you've got a reserve in your marrow of other CD20 cells of more immature that grow up and replace.

00:55:31.400 So it's not like you're really at a huge risk for untoward reactions from your CD20.

00:55:37.740 And why do those patients with marrow that's still producing CD20 positive cells not go on in a

00:55:43.520 constant state of lymphoma requiring? In other words, why is it that you can treat this and

00:55:48.760 create a durable remission?

00:55:50.420 I can guess. I don't know if anyone's done a formal study. I do think at some point in

00:55:56.180 oncology treatment, if you have a tiny amount of disease left, especially something like

00:56:01.020 lymphoma, you may take care of it yourself.

00:56:02.980 So it's just getting rid of enough of the diseased B cells until you get the load down,

00:56:08.720 the tumor load low enough that the immune system can wipe out the clone.

00:56:12.380 And honestly, if it comes back, that's the other thing that I find really interesting about antibodies.

00:56:16.760 The dogma with chemotherapy, if you are on Taxol and your tumor comes back,

00:56:22.440 I wouldn't give you Taxol again. If you're on Herceptin or Rituxan and your tumor comes back,

00:56:26.860 in a heartbeat, I'd give it to you again. Yeah. It's a very different thing than chemotherapy.

00:56:31.680 What was the price of these drugs at the time they came out? Were these the first

00:56:35.840 chemotherapeutic agents, or you kind of want to distinguish them from traditional chemo,

00:56:39.960 but were these the first oncology drugs that came with big price tags?

00:56:43.460 Probably they were. I think Taxol kind of went to that next level and then they went

00:56:48.900 to the further level compared to today's prices low. But Rituxan, I think more than Herceptin,

00:56:56.980 because people started using it more, like you'd use four times or eight times and recurring,

00:57:02.280 Rituxan sales went very high, very fast.

00:57:04.900 And about this time, we get the whole anti-VEGF story, right?

00:57:09.760 Yeah. Yeah. That was...

00:57:11.800 So Judah Folkman over at Boston Children's.

00:57:13.940 At Boston Children's. Yeah.

00:57:15.300 Yeah. I never had the chance to meet him.

00:57:16.820 Oh, you didn't meet him?

00:57:17.560 I've never met Judah. He wrote a fantastic book that I read in medical school,

00:57:22.100 poured over the book. I'm blanking on the name of it. Do you remember the book? It was his story.

00:57:25.620 Yeah. Yeah. I don't remember.

00:57:26.940 Again, a beautiful story.

00:57:28.180 I'm smiling because I had a word for Judah Folkman talks, which I heard many of. He was

00:57:33.560 a phenomenologist. He would say, this patient had this, so it must mean that. He just connected dots

00:57:40.480 all the time. I mean, some of it made no sense to me, but some of it was like, wow,

00:57:44.700 I wish I thought of that. He was just a really fun person to listen to. I used to tell this story

00:57:50.520 so many times. His thing was, the cancer can't grow larger than a BB if it doesn't have new blood

00:57:56.900 vessels. That was his thing. It stuck at a certain size. And so VEGF is the primary way,

00:58:04.500 vascular endothelial growth factor is the primary way that you grow new blood vessels if you're a

00:58:10.200 tumor cell. People went crazy about this hypothesis. It was more than the TPO that I was describing

00:58:17.540 because it was Judah Folkman and he's very compelling and very charismatic. And just because

00:58:23.220 the hypothesis really. It's logical. It's logical. It resonates. It sounds like a good thing.

00:58:28.560 You know, the saying that I love is it's the description of science as a beautiful,

00:58:33.320 compelling hypothesis slayed by an ugly fact.

00:58:39.080 That's perfect. Yeah.

00:58:41.880 Napo Ferrara, also a great character, Italian OBGYN who came to Genentech and worked in one of the labs,

00:58:49.760 made an antibody, actually same backbone as Herceptin to VEGF.

00:58:55.100 Mostly human.

00:58:56.100 Mostly human. Again, I think about 93, 94% human. And so we decided we should go after

00:59:02.820 an antibody for VEGF as our next big oncology program. I still remember, by the way,

00:59:09.660 one of the things that Gwen Fyfe, who's a great clinical oncology trials expert, 1.00

00:59:15.060 was in charge of the program. And Gwen and I talked the day before the first patient was going

00:59:19.960 to get treated with anti-VEGF. And Gwen said, my nightmare is that all the blood vessels fall

00:59:25.520 apart. We've just put that into someone's body. And I said, well, you know, we did all the talk

00:59:30.460 studies. We're like, I don't think it's going to be that bad, but we have no idea what, I mean,

00:59:34.480 it just felt... And I'm sorry, this was before the first phase one patient?

00:59:37.460 This was the first phase one patient.

00:59:38.460 This was the first phase one. So you're going very low dose. We're going to dose...

00:59:41.060 It's very low.

00:59:41.560 This is the first time it's going into a human.

00:59:43.060 It's the first human dosing of anti-VEGF. And we knew how important VEGF was.

00:59:47.800 So we were scared.

00:59:48.880 And this is also mid-90s.

00:59:50.880 Yeah. Yeah.

00:59:51.640 This is all happening when you arrive.

00:59:54.160 Yeah. It just gotten there.

00:59:55.600 I mean, what a time to be at Genentech.

00:59:57.400 Yeah. It was wild. It was wild.

00:59:59.280 Wow.

01:00:00.000 So we're into the clinic and we make progress and it's really good news and lots of studies.

01:00:06.300 And we're ambitious. We want to do a lot of different... We wanted to do lung cancer and we

01:00:10.660 want to do breast cancer and...

01:00:12.440 How are you picking the cancer to study something like this? Herceptin's obvious because you're

01:00:16.260 targeting a receptor.

01:00:17.300 Herceptin and Rituxin were easy.

01:00:18.380 Yeah. They're easy. You know what you're doing. But here, you could be targeting anything.

01:00:21.920 There is one tumor where VEGF plays a seminal role and that is renal cell carcinoma.

01:00:28.280 And yet, renal cells are really tough to study. It's just... It's not set up clinical trials-wise.

01:00:35.100 Why is that? You've got the IL-2 stuff going on where you've got 10% of people will respond

01:00:40.900 to it, but 90% won't.

01:00:42.140 No, it should be easier. It may be the sites and where the clinicians are who care for it.

01:00:47.960 It may just be that pragmatic. We kept struggling to figure out how to do a good renal cell study.

01:00:54.200 But we thought we could do a breast cancer study because we had a lot of networks of breast

01:00:58.360 cancer patients and particularly patients who weren't eligible for Herceptin because many

01:01:03.060 of them weren't. So, we wanted to do a late-stage breast cancer study because if we could help

01:01:10.380 these patients, we would find out right away. These were patients with metastatic breast cancer?

01:01:14.980 Metastatic breast cancer who had already tried...

01:01:16.780 Progressed through everything.

01:01:17.980 Everything. So, really tough high bar. So...

01:01:20.920 And the standard you're going to hold yourself to in the phase two is 50% shrinkage?

01:01:27.780 We wanted to have at least 50% shrinkage. We wanted to change time progression.

01:01:33.600 This is a great time to actually hit pause. I wanted to do this later, but I think this is

01:01:38.120 the right drug to go through two things. One, even though I've done this probably half a dozen times

01:01:43.600 on the podcast, you should never assume somebody remembers it. I want people to understand what

01:01:48.320 the difference is between a phase one, a phase two, a phase three study. Also understand what's

01:01:52.560 preclinical. It's not intuitive to people why it costs a billion dollars to get a drug to market

01:01:57.780 and why it can take a decade. And then, within that, if you could just embed enough of the details

01:02:03.040 about decisions that you can make that will make or break you. How many times has a drug failed

01:02:10.100 because the experimental design, the wrong patient selection, the wrong disease selection?

01:02:16.480 You have got to line up four pieces of Swiss cheese just right to get the pen through to hit

01:02:23.440 it. Sorry to interrupt, but let's go back to, you got Judah Folkman talking about VEGF, VEGF,

01:02:29.880 VEGF, that then turns into, well, if we made an antibody to VEGF, okay, so there's your idea.

01:02:35.840 Now, start the clock and start the dollars.

01:02:38.840 So if you start with a target, often in oncology today, we'll start with a target.

01:02:44.400 There's two things you have to start with. One is, what's the best way to turn down or turn

01:02:52.500 off that target? Is it a small molecule? Is it an antibody?

01:02:57.260 Tell folks the difference. How do you think of small molecule versus antibody? Where do we draw

01:03:00.020 the line?

01:03:00.600 So here's a really simple way I think of it helps me. Small molecules chemistry. Small molecule,

01:03:06.520 it can be, not always, a pill. A small molecule you're impacting on often pathways or enzymes or

01:03:15.920 things that happen in the cell. A large molecule, whether it's a protein or especially an antibody,

01:03:22.600 an antibody is biology. An antibody, you're trying to do something that may be immune in nature,

01:03:29.220 or you use the antibody as a delivery device. You're getting something to the cell. A company

01:03:35.340 like Genentech and many modern companies really like antibodies. I like antibodies because when

01:03:42.860 something happens, it's on target. It doesn't tend to be off target. Small molecules at chemistry

01:03:48.680 tends to have surprises in negative ways off target, like liver toxicity or kidney toxicity.

01:03:55.300 I do this through cardiovascular medicine to explain to people the difference between

01:03:58.920 a statin and a PCSK9 inhibitor. You have these two very common drugs that are used to lower

01:04:05.500 cholesterol, but a statin is a small molecule. I don't say this in an insulting way, but we use

01:04:10.880 the terminology it's dirty. It does block an enzyme, but it's got all these off target things and

01:04:16.500 your liver function gets whacked. You get insulin resistance. Some people get horrible muscle

01:04:21.400 soreness. So five to 10% of people taking this drug are going to have a side effect that prevents

01:04:26.060 them from taking the drug. I've never seen a person yet who couldn't tolerate a PCSK9 inhibitor

01:04:31.120 where you inject an antibody into them that binds to a protein and shuts it off.

01:04:38.960 That's really a good example. And the choice of molecule is driven by that. When I was first in

01:04:45.120 product development, there was this thing of, oh, you need a pill, especially for chronic

01:04:48.640 indications. You need a pill for compliance.

01:04:51.180 Right. Who would take an injection for cholesterol?

01:04:53.360 Look at obesity drugs. Turns out a lot of people would take an injection if they want to.

01:04:57.560 But once you have your selection, you need to make sure you can make it. And one of the critical

01:05:04.880 things for a biotech company, if it's a protein or an antibody is the small scale production of it

01:05:12.060 in small, they call it a mini firm. The mini firm has to resemble what is actually going to be used

01:05:18.100 because the next thing you start doing is a bunch of models. Judah Folkman giving a great talk doesn't

01:05:23.800 mean you believe that blocking VEGF will help cancer. So we do models in mice. We may do larger

01:05:31.740 animals. We do fewer animal models than we used to because they're really limited. I would rather

01:05:37.840 have a great target with good biology than an animal model, but it's still helpful. It's still helpful.

01:05:42.700 And then the critical thing is the preclinical work that you do, what FDA is going to want to

01:05:49.160 ask you, and they should, this is not them being bad. This is them being good. They're going to

01:05:53.520 want to ask you about toxicology. What's your safety plan? Based on biology of VEGF, what are

01:05:58.540 you most worried about? I'm most worried about bleeding. It's an antibody. I'm most worried about

01:06:02.260 an allergy to the antibody. Did any of the tox studies show allergy to the antibody? What are you

01:06:08.500 going to look for and how are you going to look? How often are you going to measure the patient? So

01:06:12.360 the preclinical safety plan is really important and based on what you find in toxicology.

01:06:18.060 The other thing that's essential is, and especially modern oncology, if you have a targeted therapy,

01:06:24.840 you must have a diagnostic. And that is wicked hard because you've got the therapeutic and the

01:06:31.460 diagnostic at the same time. Now, things like CD20, things like VEGF are very ubiquitous, so it's not

01:06:38.280 really targeted in the sense of HER2, where we needed a diagnostic. But if you need to have that,

01:06:45.180 we had what we called a clinical trials assay for Herceptin that wasn't to be marketed.

01:06:50.800 Did you guys have to have somebody in parallel developing a CLIA-certified assay that a pathologist

01:06:56.800 was going to use, or did you do that in-house? So we had an in-house clinical trials assay

01:07:02.800 that we used all the way through phase three. And you could quality control the hell out of it?

01:07:07.480 It was fine. It was nothing wrong with it, except it wasn't approved. So not fine. So when we went

01:07:14.120 to FDA, they said, we're not approving Herceptin until you have an approved diagnostic.

01:07:18.360 So whose responsibility is, like, how do you encourage the world to make that happen?

01:07:21.420 So we went to DACO. We went to several diagnostic companies and DACO said, we'll make you a

01:07:26.940 immunohistochemistry test for her too. And Herceptin test is made by DACO. But we had to go back

01:07:32.360 and correlate that with the clinical trials to make sure that it was the same as the clinical

01:07:37.580 trials assay. Now, why didn't you guys do that in parallel? Was the cost too great? And did you

01:07:41.320 want to de-risk the drug before you sunk the cost into that? We were too inexperienced to realize we

01:07:47.160 should have. I see. Okay. So nowadays, we're doing that in parallel. Oh, for sure, on parallel. Yeah,

01:07:51.220 for sure. It was a mistake. So how long, just to, again, go back to helping people think through

01:07:55.560 the arc of time. From the moment you guys hit a go decision on, we want to do this,

01:08:00.380 we want to pursue this path. How long until you file the IND? Oh, gosh, it could be years. It could

01:08:06.560 be two, three years because you're doing animal models. Maybe tell folks what the IND is so they

01:08:11.020 understand why that's an important milestone. So the investigational new drug is asking the Food

01:08:16.180 and Drug Administration permission to ship an unapproved drug across state lines. If you and I wanted to do

01:08:23.460 something in Austin, we could actually do it, which is sort of weird when you think about it.

01:08:27.760 But most people don't really want to do that. So we're going to do the Peter Sue drug. It's going

01:08:31.540 to be amazing. We're going to set the lab up right over there. But the moment we want to run a clinical

01:08:36.000 trial and ship it and get it out of the state, you got to have the IND. So the investigational new drug

01:08:41.520 is the request. And what happens is that you take all this information I've been talking about,

01:08:47.020 that you know you have a molecule, you trust the way you're producing the molecule,

01:08:52.480 you understand the biology enough, you have a safety plan, and you have a phase one protocol.

01:08:58.160 So phase one has one purpose. We're all greedy. I've been there. It is only for safety. Phase one

01:09:07.220 is, is it safe to give humans this molecule? Is it safe to give it once? Is it safe to give it

01:09:12.440 multiple times? And there's an art to knowing where to start the dose because it's an escalating

01:09:17.360 dose trial. That's right. But you're extrapolating from what you learned about toxicity in a totally

01:09:23.020 different organism that never translates one-to-one to the organism of choice.

01:09:27.520 It's absolutely true. And it's not uncommon. And you see people all the time backing up on the dose

01:09:31.960 thinking, oh, that was more than we needed or more than we wanted. But phase one with a good

01:09:37.920 preclinical package, a good IND, phase one should be uneventful. And because we are greedy in oncology,

01:09:44.660 we always look to see if anybody responds just because that's what we do. But phase one often,

01:09:50.320 to be fair, has some really tough patients who are trying something and have tried a lot of other

01:09:55.760 things. So the patient population can be tough to find any efficacy in. So phase one, I always think

01:10:01.320 of phase one as it might be a year that you're in phase one if you're doing a really good job.

01:10:06.200 And typical cost given the relatively low numbers of patients?

01:10:09.400 Oh, gosh. In the tens of millions?

01:10:12.440 Tens of millions. Yeah.

01:10:13.520 Yeah. Tens of millions. And then you get into the 20s and 30s and 40s of millions with the phase two,

01:10:19.320 depending on how big your phase two is. And phase two, I think that's where people can

01:10:24.160 use their intellect, I think, in many ways. Phase two, you start to look at what's the right

01:10:30.500 dose and schedule. Very, very important to get the right dose and schedule.

01:10:35.240 And what's the right outcome? What's the right patient?

01:10:37.140 Who do you want to treat? And really what phase two is supposed to do, with one exception,

01:10:43.720 phase two is supposed to get you ready for phase three. You've got a dose, you've got a schedule,

01:10:48.080 you've got a patient selection criteria, and you've got a hypothesis of where this is going

01:10:52.600 to be a drug. The exception in oncology is sometimes you want to get an approval off phase two.

01:10:57.980 When we tested anti-VEGF in breast cancer in phase two, we wanted to use that as an approvable

01:11:04.420 study. Because we would go in and say, look, it can be a contingent approval, but these patients

01:11:08.720 have nothing else to do. And so I think that can happen, especially targeted therapy, where you've

01:11:14.420 got the perfect target and FDA is feeling good about it too. That can be a phase two study. But

01:11:20.020 most of the time you're getting ready for phase three.

01:11:22.300 So where were you guys with anti-VEGF in phase two? You're at breast cancer. And did you do colon?

01:11:28.120 We did colon, but not the kind of study that I just mentioned for approval. We did a traditional

01:11:33.680 phase two in colon. So what happened is the breast cancer study failed. And I was so disappointed.

01:11:39.520 I was so hoping that that would work. I still remember that day. For me, it was like, oh,

01:11:45.140 we need more better drugs for breast cancer. Because I often heard from people when they're the three

01:11:50.420 out of four patients, suerceptin wasn't for them. If you looked at your stock that day,

01:11:55.220 it also looked really bad because all the hype about Avastin was there. But in colon cancer,

01:12:02.700 we had a phase two, got ready on the dose and schedule. And then we went to a phase three

01:12:08.580 study in colon cancer, much more traditional, just plus minus Avastin.

01:12:12.560 5FU and the usual suspects.

01:12:14.700 5FU and the usual suspects, plus minus Avastin. And that succeeded.

01:12:18.280 That was a stage four?

01:12:19.280 Only stage four.

01:12:20.300 This is a median survival study. You're not doing overall survival, correct?

01:12:23.180 That was a median survival study. Actually, I don't remember all the details of that one.

01:12:27.580 I feel like it was eight more months of median survival. Does that sound about right?

01:12:30.920 Probably right. It was the first new thing in colon cancer for a while too. So people were

01:12:35.780 pretty excited.

01:12:36.480 Now, at this point, I'm in medical school, just down the street. I'm at Stanford. And I remember we

01:12:42.440 had a big discussion about this. I'm in my first year of medical school. And the discussion we had in

01:12:46.360 class was, I think at the time, Avastin was $100,000 for the treatment. Extends median survival

01:12:52.180 by whatever, but I think it was eight months. The UK said no. The NHS said, we are not paying for this

01:12:58.080 because at the time, the NHS had this $100,000 quality adjusted life year hurdle,

01:13:04.760 which is understandable, right? That's how you throttle supply side economics. They said, look,

01:13:08.900 we can't pay for a drug. We can't pay more than $100,000 for an incremental year of quality

01:13:13.740 adjusted life year. This is only giving you eight months. That's why I know it was less than 12.

01:13:17.800 And so the NHS flatly said, we're not paying for this. And I do believe people in the UK could pay

01:13:22.480 out of pocket for it.

01:13:23.580 You can get it out of pocket, but not through the National Health Service.

01:13:25.720 That's right. People in Canada could not because you couldn't have private insurance in Canada,

01:13:29.820 though you could come to the US for treatment. So of course, this just became a great topic of

01:13:34.120 discussion for med school freshmen. What was your thought at the time of, have we moved the needle

01:13:40.520 enough? How do we think about the economics of this?

01:13:43.080 So I had a lot of different reactions. First of all, with Avastin, it was the first time I remember

01:13:50.040 reading, and I think it was one of those curtain raiser things in the Wall Street Journal for the

01:13:54.980 breast cancer study. The headline was, Avastin might help breast cancer patients, but can they

01:14:00.780 afford to take it? I was shocked that it was the first time I had read, and as long as I had been

01:14:06.620 at Genentech, that somebody couldn't afford one of our drugs. That instead of saying, oh, isn't that cool,

01:14:12.320 Britoxan, Herceptin, Avastin, it was too much money. And that felt really important to me and

01:14:19.460 really not good. We had as a company a philosophy that no patient should go without any of our drugs

01:14:25.800 because of an inability to pay. So we had a bunch of patient, what do you call those, patient support

01:14:30.900 programs or whatever. So we had a bunch of different things in place. So I knew we had those programs,

01:14:35.880 but that doesn't help the patients in the UK, and it doesn't help the overall cost because we're

01:14:40.400 actually supplementing, but the cost is still really high. And we started to have a lot more

01:14:45.560 discussion at the executive committee about the price and how we would think about it and how we

01:14:50.200 would price the drugs because that was, like I say, that was not good.

01:14:54.700 Did you go straight from Genentech to being the chancellor at UCSF?

01:14:57.920 Just because I want to stay with the story, I want to continue the arc of your career.

01:15:02.360 We're at the halfway point, right? We're one third of the way into-

01:15:05.740 Early days.

01:15:06.060 Yeah, yeah, yeah.

01:15:06.660 We're in the 90s.

01:15:07.760 How did you, given you are truly on the cusp of what is happening in oncology and biotechnology,

01:15:15.380 and now the same institution that said, you can't come back here to have a clinical appointment after

01:15:21.780 saving the people of Uganda is now offering you the highest post, essentially, outside of a

01:15:27.880 provost, I'm guessing, right? I don't even know where the chance-

01:15:30.860 It is the highest post.

01:15:31.180 It is the highest post. Okay, so that's kind of remarkable. And does that just speak to

01:15:35.740 what they saw as the vision of that institution, which was few people have learned what Sue has

01:15:41.120 learned in the last 10 years, and we want that type of leadership here?

01:15:45.880 So first of all, on the Genentech side of things, Roche bought us.

01:15:49.540 That's right. That was 99?

01:15:52.020 That was 2009.

01:15:55.000 Oh, 2009. Oh my God. Okay, I'm off by a whole decade. I thought that was sooner.

01:15:59.540 2009, which was not what we wanted. It was a-

01:16:02.620 It's a hostile takeover.

01:16:03.400 A hostile takeover, yeah.

01:16:04.560 They call it in the business world a squeeze out. They squeezed us out. So I knew I was going

01:16:09.280 to leave. I knew I was going to do something different. And UCSF, I had been really close

01:16:15.140 with the chief of medicine, who was my chief of medicine when I was an intern, Holly Smith. Do

01:16:21.520 you know Holly Smith?

01:16:22.180 No, I don't.

01:16:22.880 So Holly Smith was a South Carolinian, Harvard-trained, amazing person. And between him and Bill

01:16:30.860 Rudder, who founded Chiron, who's a biochemist, they decided a long time ago that UCSF should

01:16:38.100 not be a backwater medical school and should be a serious medical school. So Holly on the

01:16:43.580 clinical front and Bill on the scientific front just decided they would start recruiting people

01:16:48.920 to have a great institution, like a pretty amazing thing.

01:16:52.320 And that was starting in the 80s?

01:16:54.020 It was probably 80s, 70s, 80s. I was still friends with Holly. I just think there's so

01:17:01.980 many wonderful things about Holly that I admired that I was still friends with him. So Holly called

01:17:07.400 me and said, Mike Bishop is stepping down. Of Bishop and Varmus, that Mike Bishop, yeah.

01:17:12.840 Of Nobel fame.

01:17:14.600 Yeah. And you should be chancellor. Of course, I said to Holly, I'm not a professor. And I said,

01:17:20.260 well, if they're interested in me, I'd be open to talking to them. They were. And I went through

01:17:25.820 the interview process.

01:17:27.380 I'm sure that when you're going through that, sorry to interrupt, because you're now interviewing

01:17:30.560 with the board of trustees.

01:17:31.800 They have a search committee.

01:17:33.020 Yeah.

01:17:33.300 Yeah.

01:17:33.540 And so they must be asking you to present a vision. They're not interviewing you to make

01:17:37.960 sure you know how to use PowerPoint.

01:17:39.480 Yeah.

01:17:39.980 Do you remember what the vision is that you presented to them?

01:17:44.080 What I do remember, and I remember, if you remember 2009, was just...

01:17:50.020 Post-apocalypse.

01:17:51.360 Horrible recession and California being particularly in a bad place. So I just talked to them about

01:17:58.740 how I think about managing people, how I think about making sure that you use whatever assets

01:18:04.040 you have maximally. I admitted that I had never done fundraising, but I had done a lot of work

01:18:09.440 with Wall Street and I could talk.

01:18:12.060 You were the president. What was your title at Genentech before you left?

01:18:14.860 Yeah.

01:18:14.980 And I did a massive amount of investor relations because Art didn't like traveling or talking.

01:18:22.060 My kind of guy.

01:18:25.080 Mine too. No, we were a good partnership, but I also said, look, if you think about running

01:18:30.860 the faculty meeting, I'm probably not your guy, but that's what the provost does. I think they

01:18:34.980 really thought they wanted somebody who could work on the business aspects of the campus.

01:18:39.560 Because we needed to finish Mission Bay.

01:18:41.420 You're really the CEO of the system.

01:18:43.900 Yeah.

01:18:44.280 Yeah. And you manage the hospital CEO and it's a big hospital system. There's no undergrads at

01:18:49.060 UCSF. So that was what they were looking for. And I thought, well, why not?

01:18:55.300 Were you nervous? I mean, when they called you and said, you've got the job, was there a moment

01:18:59.160 where you thought, have I bit off more than I can chew? This is a huge responsibility.

01:19:03.540 I was really concerned. I was really concerned about it. And I also realized because some of

01:19:10.260 the faculty were pretty negative when I first started.

01:19:13.480 Because you're an outsider, you were an alum, but you weren't coming up through the ranks

01:19:16.880 as the CEO of the hospital or something.

01:19:19.240 I actually think they were just as nervous as I was about money. They weren't convinced that I

01:19:24.220 knew how to get them money. Because, you know, if you're running a program, you need money.

01:19:27.880 It's the mother's milk of doing science.

01:19:29.680 So tell me about the budget of UCSF. Because it's a state school, presumably California provides

01:19:35.460 what fraction of it?

01:19:36.960 Almost nothing.

01:19:38.020 So what's the benefit of being a UC?

01:19:40.240 The brand. There's a curriculum thing. There's some things like that.

01:19:43.400 So in that sense, you and Stanford aren't that far apart.

01:19:45.880 Not that different. I mean, the most important things, you've got clinical income,

01:19:49.680 you've got NIH-driven income, right? You've got other grant income, and you have philanthropy.

01:19:54.860 Show me the P&L on those things there. So NIH is bringing in how much?

01:19:58.660 Oh, gosh.

01:19:59.560 Percentage-wise, roughly.

01:20:01.020 Of the money that you use every day, there's this discussion of overheads now. It probably

01:20:05.580 gets up to a third.

01:20:07.760 A third of the revenue for general operations is coming out of the NIH overhead.

01:20:13.200 Probably, yeah.

01:20:13.900 And then clinical revenue.

01:20:15.560 Clinical is a lot, but a lot of it goes back to the hospital. You know, it's a not-for-profit.

01:20:19.440 They spend it on the hospital.

01:20:20.920 Okay. And then philanthropy is some direct, some into the endowment where you're living off the interest.

01:20:28.120 Yeah.

01:20:28.580 And that's basically what your revenue streams are, those four things.

01:20:31.760 Yeah, that's the revenue stream. And then the tuition is tiny.

01:20:34.200 Well, especially because you don't have undergrads, right?

01:20:35.720 Right, right. Because the number of students is really low. And the really good news at UCSF...

01:20:40.580 I didn't even apply to UCSF, by the way, because, I mean, I was not in California when I was applying

01:20:44.640 to medical school. I was told, well, such a great medical school that I was like,

01:20:48.500 there's no way I'm going to get in as a non-Californian. So I didn't even apply.

01:20:53.180 So the funny thing for me is...

01:20:54.640 But you did, and you got in.

01:20:55.680 I did. Well, there's a story there. We talked about me being at University of Nevada. What

01:21:00.440 we didn't talk about is my first year at University of Nevada, my youngest sister was born.

01:21:06.660 Wow.

01:21:06.900 I lived at home and helped my mom in Reno.

01:21:11.280 Because there's like seven of you, right?

01:21:12.700 There's seven of us.

01:21:13.420 Yeah.

01:21:13.660 Yeah. And number seven was born when I was a freshman. It was kind of crazy. But when I

01:21:19.240 went to medical school, I used to have like this sign on the stairs, be quiet. I'm studying

01:21:24.340 down here. It's like probably a giant pain. But I wanted to go to San Francisco. My dad was

01:21:30.920 born and raised in San Francisco. My grandma lived in San Francisco. So even though I was still

01:21:34.900 pretty young in terms of ever living outside the home, I knew San Francisco. I wanted to

01:21:40.680 go. But UCSF had never taken a University of Nevada student because we were a pretty new

01:21:46.240 medical school. I was only in the second four-year class.

01:21:48.940 Oh, wow.

01:21:49.500 Yeah. And I was excited like my head would blow off. Like I wanted to go to UCSF so big

01:21:57.300 time. So I'm going back to UCSF. And the really good news, what saved me as chancellor is philanthropy.

01:22:06.320 It turns out that we needed people to care about the mission and the projects at UCSF right

01:22:14.100 at the same time as the Mark Zuckerbergs of the world and the venture capitalists of the

01:22:19.600 world. And a lot of people had come into a lot of money. Even though the overall economy

01:22:24.200 was bad, it was coming back. And we had some spectacular successes. And my successor, Sam

01:22:31.420 Hogood, continues to have that kind of success. And people are just really generous.

01:22:35.760 Americans are, hands down, the most generous people in the world. I think that's a demonstrable

01:22:41.480 fact.

01:22:41.960 Yeah. I'm big on New Year's resolutions, among other things. One of my New Year's resolutions,

01:22:47.800 I have a mini list. One thing that's always on the list is be more generous. I'm no Mark

01:22:53.300 Zuckerberg, but I can be generous in other ways.

01:22:55.860 I'm sure that you must deserve some of the credit for that. I don't think it's just that

01:22:59.140 a bunch of people in the Bay Area came into money at that period of time. What was the approach

01:23:04.680 you took towards philanthropy? And how did you reach donors that maybe previously hadn't

01:23:10.120 been involved in UCSF? Because again, one of the things that's working against you is you don't

01:23:14.560 have an undergraduate. So Stanford has a big advantage over you in which you've got a lot of

01:23:20.160 people that are coming through doing engineering degrees, doing CS degrees, who are going on to

01:23:25.180 create enormous enterprises. Anyone who's an alum of UCSF went to graduate school there. There's

01:23:31.760 no business school. There's no law school. So you're missing out on a lot of this.

01:23:36.040 Don't forget the hospital. The most important donor base is grateful patients or people who

01:23:41.120 love science, technology. I hired John Ford, who had retired as the head of Stanford's fundraising,

01:23:49.000 moved up to the Northeast and he was my head of development. And I talked to John and I said,

01:23:54.320 how do you do this? Teach me how to be a good fundraiser. And he talked about, tell people your hopes

01:23:59.860 and dreams. Tell people what you're excited about and ask them what they're excited about and see if

01:24:05.140 there's a match. And I think that was really important. And then I also think that because

01:24:10.400 I had been at Genentech and I was sort of gregarious and knew a lot of people and people knew that I had

01:24:15.600 decent business savvy. I wasn't going to waste their money. We were very committed to use the money

01:24:21.040 wisely, especially in the hard times and do special things at UCSF. So I was surprised. I sort of worried

01:24:29.480 that I would be sad if people said no, that it would be weird, especially if I knew them well.

01:24:35.140 So I would get myself psyched up for the beginning of it. And then by the end, I'd be like, oh, that's

01:24:39.660 fine. Next time if you're in town, let's talk again and maybe it'll change or whatever. But it was

01:24:44.980 actually fun. I got to talk to and meet a lot of great people.

01:24:48.380 What percentage of your time was spent externally versus internally?

01:24:54.200 Probably 40% externally. A lot external.

01:24:57.080 And what was the internal focus then?

01:24:58.860 Working with the team.

01:25:00.160 So who were your direct reports? The provost?

01:25:01.640 Provost, CEO of the hospital, lawyer, all the deans. That was really important. And then part of it was

01:25:10.180 monthly we met with all the chancellors, with the president of the university. It was Mark Udoff at

01:25:14.800 the time and then he stepped down after a while. But the chancellors meetings were funny because

01:25:20.060 they all had undergrads. I always felt like I was squirming, like are we done yet? I thought it was

01:25:26.640 just really important. My favorite thing, every Friday, lunch, Mission Bay, they had a science talk

01:25:33.480 and they'd have some pizza, Chinese food, something. And you'd look around the room and it'd be like

01:25:40.400 Bruce Alberts, Liz Blackburn. There'd be four Nobel Prize winners in this little cramped room

01:25:46.620 listening to science.

01:25:48.780 This would be something you did as a scientist, not necessarily as a chance, like the chancellors

01:25:53.700 weren't doing this all the time.

01:25:54.760 No, I'd just go over on the shuttle bus, eat a slice of pizza and enjoy.

01:25:58.160 Yeah, amazing.

01:25:58.880 It was really good.

01:26:00.200 So then let's get to the next chapter. What all of a sudden in 2013, 2014 leads to the next

01:26:05.640 transition to being the CEO of the Bill and Melinda Gates Foundation?

01:26:08.800 Well, to my surprise, I think I got an email or I think it was an email from Melinda.

01:26:15.920 Did I have time to talk? And UCSF throws a big event every year, this kind of friend-making,

01:26:22.780 fundraising, everything, and we give out awards, recognition to people whose work we respect a lot.

01:26:28.800 So I had invited Melinda the year before and thought, she'll never come. You invite people,

01:26:34.600 the throwaway invite. And she accepted and came, actually came with her mom and dad even better.

01:26:40.020 I thought that was nice. So she sent me an email and she said that she and Bill wanted me to look

01:26:48.460 at being the next CEO of the Gates Foundation. And I was surprised. I had not expected that. And

01:26:55.180 I started having discussions with them. It was actually funny. I went up to Seattle and they

01:27:01.960 were having all this hush-hush. You know, this was very cloak and dagger. So I went to their house

01:27:07.500 because I had had a meeting with Melinda and I needed to meet with Bill. And it was Halloween.

01:27:12.200 The kids are coming and going. It was kind of crazy. And so I had talked to him and I talked

01:27:18.700 to my husband. You know, my husband worked at the Gates Foundation. He led the HIV and TB programs

01:27:23.940 about five years before.

01:27:26.060 Was he still there?

01:27:27.060 No, no, you had gone. He was commuting to Seattle, which was dreadful.

01:27:30.560 But he knew the Gates Foundation. So he and I were talking about this and I was like, oh God,

01:27:34.360 you know, I've only been at UCSF five years. I just found where the bathrooms are, you know,

01:27:39.120 that kind of thing. And it was going well. I was happy with that. They asked for a teleconference

01:27:44.260 and they got on the phone and especially Bill made this big pitch that a lot of people could

01:27:52.220 do the UCSF job with all due respect, as Bill would put it. But I was the only person who they

01:27:58.320 both wanted and who could do this job. I was perfect for the job and it's really important

01:28:03.100 for the world and I needed to do it.

01:28:04.620 I assume that the rationale for that was obviously their focus is on global health and you have the

01:28:14.020 background in the clinical side, the research side, the epidemiologic side, the management side.

01:28:20.200 So there's kind of those are four legs of a chair. Were there other things that I'm missing that they

01:28:25.240 felt were kind of essential pillars?

01:28:26.980 I think it was less obvious then, but I think now they had started to kind of have disparate views of

01:28:32.980 how the foundation should operate. Melinda has been really all over women's issues, all over. 0.77

01:28:40.440 And Bill would do another run at polio. You know, it's like the goal broadly that all lives have

01:28:47.600 equal value, which by the way, I think is a wonderful thing. They share, but they come at it

01:28:52.440 from different ways. And so I think that the thing that resonated for me is that I could see both

01:28:57.920 those points of view. But those points of view don't strike me as mutually exclusive for an

01:29:02.380 organization with enormous resources.

01:29:04.580 Yes and no. It's one thing to have enormous resources. It's another to know where one of

01:29:10.080 the most important assets they have is the time and energy of Bill and Melinda. They actually show up,

01:29:17.120 things happen.

01:29:18.560 So how did you weigh this decision?

01:29:21.240 I thought that I could add value. I thought I would learn a lot. And I did think that UCSF

01:29:27.160 would be fine without me. I felt like we were back on our feet financially. I thought that Sam

01:29:32.140 Hogood, who was the dean of the School of Medicine, I had a ton of respect for him and thought he was

01:29:38.660 the obvious person to take my place and that it would be okay.

01:29:42.540 Were they surprised? Did they try to talk you out of this?

01:29:45.840 I don't think they did. I actually think they had a lot of respect for the Gates Foundation and

01:29:49.880 thought, oh, well, that's a cool job. At least the way they showed up with me. Maybe when I wasn't

01:29:54.860 there, they did.

01:29:55.980 Okay. So you head up to Seattle now.

01:29:57.440 Yes.

01:29:58.100 When you show up to the foundation, how many employees are there? What does it look like?

01:30:01.580 It's a not-for-profit, but does it run like Microsoft? I mean, how does it operate?

01:30:05.320 It's a couple thousand. It's a big foundation, big building, big foundation with people all over

01:30:10.900 the world now. There was a lot I wasn't surprised by, like the global health stuff. I knew what they

01:30:15.460 were doing and I thought it was interesting and great. And the challenge for me was Bill's

01:30:22.300 endless appetite for things like learning things, driving things, funding things, and me feeling

01:30:31.700 like I could get my hands around a strategic plan. It was a little bit like, okay, this staff would be

01:30:38.040 like, Bill's going to love this. Let's present that. You know, it was that kind of feeling and lots

01:30:41.960 of money. So I kept trying to get my hands around like, okay, what should we do so it's just a little

01:30:48.600 more orderly and we get a great return on our investment. That was the most important focus.

01:30:55.020 I feel good about that. The funny thing was I sent the finance team to Genentech and we had this really

01:31:02.080 great portfolio management process that we put in place when I was there and they still use it

01:31:07.080 apparently. And because I recommended to Bill, we just have a portfolio management process. Pretty

01:31:12.160 simple. Everybody knows how you make decisions, what money's up, what we'll do, and we can use that

01:31:18.120 here. It doesn't need to be bureaucratic. Bill said, we don't need it. It's all in my head.

01:31:23.460 I remember that conversation and I thought, if it wasn't you, I would think that was a smart-ass thing

01:31:28.600 to say, but I actually think you're just being honest. So I encouraged him to understand that just

01:31:34.920 because it was in his head didn't mean that the rest of us were there. We had a little more ability

01:31:40.960 to make things orderly, I'd say. It was a wild ride. It was six years of a wild ride.

01:31:48.080 What was the most difficult thing for you to impact that you wanted to change? Meaning,

01:31:54.500 was there a global initiative that you wanted to get your hands around that you just couldn't do

01:31:58.340 organizationally or technically or what were the challenges?

01:32:01.560 I would say far the opposite. The things that I felt like I knew about, I felt like really good

01:32:07.880 about. The TB stuff, there's a HIV cure program now that I'm really psyched about. Technically,

01:32:14.060 I felt really good. Probably the hardest thing for me was the people side of things. I have a very

01:32:20.540 strong sense of how people should treat each other and the competencies that managers should have.

01:32:28.940 And I'm not willing to move on that because you're a technical expert. And I found that if you do move

01:32:36.940 on that, it just changes the culture. And I struggle with that.

01:32:41.240 Say a little bit more on that. Is that because in the not-for-profit space, you have a different

01:32:46.320 talent pool than you do at Genentech?

01:32:48.660 No, I think it's because Bill really likes technical experts. And if he likes the technical

01:32:54.560 expert, he doesn't want the CEO to come and say anything. But yes, they are very smart.

01:33:00.860 What did the org chart look like? So I assume Bill and Melinda are co-chairs?

01:33:04.500 They were, yeah.

01:33:05.520 And then as the CEO, who are your direct reports? Is it organized by a bunch of GMs in different

01:33:11.400 programs?

01:33:11.960 Yeah.

01:33:12.080 So there's a TB person, an HIV person, a polio person, a malaria person?

01:33:16.720 No, there's a global health person. There's a global development person. There's an ag person.

01:33:21.080 So under global health, you then have sub-

01:33:23.720 Then you have the subs. Yeah. 0.86

01:33:25.520 Global health is a big job, obviously.

01:33:27.480 It's a very big job.

01:33:28.260 That's the biggest-

01:33:28.600 Yeah. It's the biggest P&L. And then U.S. education is a pretty big job too.

01:33:33.540 Oh, I don't even realize. I'm not as familiar with the portfolio.

01:33:35.960 Yeah. The one that has been ag is now ag, financial services for the poor. So it's a

01:33:42.620 pretty broad group.

01:33:44.240 What's the annual budget?

01:33:45.380 $8 billion.

01:33:46.520 Wow.

01:33:47.400 So much money. It's amazing.

01:33:49.900 Yeah. So what are things that you could not have done there in that role had you not had

01:33:56.020 the leadership roles at Genentech and UCSF?

01:33:59.460 Oh, gosh. I think more the people side of things. I remember there was an employee who

01:34:06.600 was really struggling at Genentech. And her boss, I was his boss. And he kept talking to

01:34:12.560 me about how she was struggling, how she was struggling. Could we do this? Did we need to

01:34:16.240 give her fewer reports, more reports, make her job harder, make her job? We couldn't figure 1.00

01:34:21.140 it out. Couldn't figure it out.

01:34:22.160 Performance was struggling, you mean?

01:34:23.500 Performance was struggling. And just not acting like she had been. We just couldn't figure it

01:34:27.920 out.

01:34:28.780 And finally, one day she said, oh, I'm getting a divorce. And after a little while, things

01:34:34.380 got better. And I thought, you know, not everything's work. Not everything's work. So I think as a

01:34:40.740 manager, I really care about people thriving at work. I really care about it. And when I

01:34:46.440 went to Gates Foundation, I think I understood better, given Genentech and UCSF, that a very

01:34:53.340 important principle, work never fills in for home, ever. It never makes up for a bad

01:34:58.540 home. So if somebody needs a timeout, I always think, how can I improve work? And sometimes

01:35:04.140 it's good to just understand that that's not always the case. Especially if you're working

01:35:09.700 in global health or global development. You might be in South Africa. You might be in China.

01:35:15.520 It's rough. So just thinking a little bit about how people can show up in ways, it's $8 billion.

01:35:22.580 How do they maximize the benefit of that $8 billion? And what can I do to enable that?

01:35:27.800 Are you basically only able to affect that through your interaction with your direct reports and just

01:35:34.360 hope that that's the infusion of culture that then trickles down? Because it's hard to go two

01:35:40.080 levels below your management. And yet the people who probably need this compassion the most are

01:35:46.900 people you're not even going to meet.

01:35:48.840 Yes and no. One thing about traveling a lot, you have big events or things like that, is you meet

01:35:54.480 people on trips? And that's different people throughout the organization. So I think there

01:35:59.060 are opportunities. I also set up meetings when people would have a grant that needed to be signed.

01:36:04.680 The business process was it would show up on my computer. So I changed the business process so I

01:36:09.160 got the group that could fit around the table in my office and we would talk about the grant.

01:36:13.640 So I could interact with more people that weren't my reports, which I really liked.

01:36:18.740 So I do think it is mostly through your reports, but I think there are ways that at a senior level

01:36:26.420 you can interact with people culture-wise.

01:36:29.020 Yeah. The word culture, it's very misunderstood. When you think about the culture that you wanted

01:36:35.060 to bring to the Gates Foundation, I'd like to understand what that was and how successful you

01:36:39.300 think you were able to be. And I say that because you were in an organization where you also had very

01:36:44.280 powerful other present people whose culture was also a part of the organization.

01:36:49.160 So for me, I define culture in a really specific way. That when you come to work,

01:36:56.340 you feel like the atmosphere, the surround sound brings out the best in you. And that you have some

01:37:03.480 ownership of tweaking it if it doesn't. So that's something that you feel like you can control.

01:37:09.120 Because if you're in thousands of people or hundreds, thousands, tens of thousands of people,

01:37:14.700 and you're the CEO, you're not going to do that. But that you set up that culture.

01:37:19.280 One of my favorite stories from Genentech was being at a product development meeting that my

01:37:27.160 successor as medical officer, Hal Barron, was running. And Art and I both attended just because

01:37:33.320 we loved it and we wanted to be there, but we weren't decision makers. We were just attendees.

01:37:37.180 Which is kind of odd. The person who's the decision maker is sitting in the presence

01:37:40.760 of the two most senior people in the company, leaving it to him to make the decision.

01:37:45.340 We're leaving it to him. Yeah. But in this case, there was someone who the discussion was about

01:37:49.900 her septum and how well the test to find who was hurt who positive performed. And if you got more

01:37:55.660 patients, you would get more commercial, but you would have patients who wouldn't benefit. And someone

01:38:00.980 who I won't name said, but if we have a test that does like this, we can get more money.

01:38:05.780 And as if in unison, Art and I both rose up from our chairs and said, we never do that. We don't do

01:38:13.120 that here. Done. Everyone kind of sat down and we weren't the decision makers. But see, that's

01:38:18.660 culture. If you wondered, it's right where the decisions are being made and everything else.

01:38:24.040 That for me is culture. The other one that because I'm in so many meetings and wasn't so many meetings

01:38:29.520 at Gates Foundation, I had a practice. I would sit, Bill would be there or Melinda would be there,

01:38:36.620 but often Bill and me there and you're presenting. And Bill's peppering you with questions, some of

01:38:41.940 them very tough, in a very tough way. I would look at you. You got this. I can't tell you how many

01:38:47.700 people, I actually didn't even know I did it, that I would nod. You were the coach. Smile, coach.

01:38:53.780 I would also stall. Hey, hold on a minute, Bill. I think he's just getting ready to answer that

01:38:58.640 question. You're talking over him right now. In a nice way, not confrontational. That for me is

01:39:04.220 culture. I want you to succeed. I want you to know I want you to succeed. It's the guy who runs the

01:39:10.200 foundation. It's the two people who are the chairs. So it's going to be scary. That's why I think that

01:39:15.840 bringing out the best in people and giving people agency to do that on their own means that if

01:39:22.980 somebody sees Art Levinson say, that's not the way we do it here, they'll go down the hall in a

01:39:28.540 different meeting and say, you know, I heard Art say, I think that's really powerful. That's really

01:39:33.140 powerful. So this brings us up to 2020. And were you at the foundation when COVID hit?

01:39:40.620 I had announced that I was leaving and literally packing the house as COVID hit. Actually a few

01:39:48.180 miles from us, the first case in Washington state nursing home.

01:39:52.240 Yeah. Interesting. Let's talk a little bit about COVID. So I've talked before about this idea of the

01:39:58.120 difference between science and advocacy, and I still haven't really wrapped my mind fully around it

01:40:03.820 other than kind of a sense of lost opportunity with COVID. What do I mean by that? Well, on the one

01:40:10.100 level, there was so many pretty incredible things that happened with respect to the speed with which

01:40:16.060 a vaccine could be developed that really made a difference in terms of mortality for a subset of

01:40:21.180 the population. But a lot of that's overshadowed today by the lingering doubts, the lingering suspicions,

01:40:28.940 the mistakes that were frankly made. And my fear is I'm not convinced we're better off today

01:40:36.220 in terms of preparedness for a pandemic than we were in 2019, which seems like an unimaginable

01:40:43.400 statement given what we've been through. Do you think I'm too pessimistic? How do you feel?

01:40:49.640 I do not think you're too pessimistic. I am absolutely horrified. Horrified. It's shocking to

01:40:56.280 me that the narrative is in the place it is today. And I'm honestly still processing how we got here

01:41:05.380 from there. It is a really bad place. And I think you're right. I think if it happened again today,

01:41:12.120 it would be the way it was with worst of COVID, but even worse. Because trust and the need to have

01:41:21.360 sides and winning and losing, I don't remember health and medicine being winning and losing and

01:41:29.200 sides as I've been in this business for 40 years. It's just weird. I don't get it.

01:41:35.680 Yep. I concur with all of that. And I do wonder what it will take to restore confidence. Look,

01:41:44.240 we could sit here and talk about mistakes. It might be that the medical community and the scientific

01:41:49.340 community need to be more vocal about acknowledging mistakes. And I do think an enormous mistake,

01:41:56.300 though it's understandable to me why it happened because so much was happening so fast. But I

01:42:01.340 believe deep down it was an enormous mistake to be the head of science, to be the head of advocacy.

01:42:06.740 I think having Dr. Fauci as being both of those hats was a cataclysmic error. And it's not about him.

01:42:14.380 No human can do that. A scientist has to be an impartial observer of fact who is happy to change

01:42:24.760 his or her mind in the presence of new information with no attachment to what has been said in the

01:42:30.840 past. An advocate has to be driving policy and action. And sometimes they have to settle for the

01:42:38.580 best you can do any port in a storm. When you put those two hats on the same people,

01:42:44.700 I worry that you lose all trust. I do wish the medical community could have an open and honest

01:42:49.700 discussion about that. I would say that not if, but when we will have another pandemic. There's

01:42:55.660 zero doubt in my mind. Bird flu is working hard on it right now.

01:42:58.720 We will absolutely have another pandemic. I hope it is decades from now, but we will.

01:43:03.120 I hope somebody will remember that lesson and say, we want our chief communicator of the state

01:43:10.220 of the science to be completely uninvolved in telling the public what to do, simply there to

01:43:17.900 report what we know today. Today, we think masks work. You know what? We just did a study and we

01:43:23.160 realize they don't work worth a lick. Today, we believe vaccines prevent transmission. We just did a

01:43:29.620 follow-up survey. They don't prevent transmission. It's okay. It's okay. I think that's a very

01:43:34.640 forgivable position. I think the public would welcome.

01:43:37.980 Look, I just told you about bone marrow transplants for breast cancer. If you tell people, look,

01:43:43.720 here's what we thought. We thought harder treatment was better for people. It's now proven that it's

01:43:48.480 not. Science changes. People know that, but you're right. I think that being honest and open when it

01:43:54.500 changes and how it changes matters a lot, it really does. I also think you didn't say, but I would add

01:44:01.540 to your recommendation, which I think is a really smart one, the pace of communication, the social

01:44:09.100 media and misinformation or just stuff gets out there really fast. And having something slow doesn't

01:44:19.060 keep up. I don't have a solution to that. I mean, the great example, which is a very good example,

01:44:24.540 and I don't know the solution is in May of 2020, if you suggested that this came out of the Wuhan lab,

01:44:34.900 I mean, you were kicked off social media, you were kicked off YouTube, you were in the doghouse.

01:44:40.800 That was misinformation. Well, I think almost any observer today would agree that that was actually

01:44:45.220 information. But where do you draw the line? I don't have an insight. This is so far above my

01:44:50.080 pay grade. Yeah. I don't think it's a matter of kicking people off because actually I think you

01:44:54.300 enhance that and you may be wrong, but being part of the dialogue, I'll give you an example that I've

01:45:00.520 been reading the last couple of weeks. Ivermectin for cancer. Actually, I'm glad you brought that up.

01:45:05.920 I wanted to have a discussion about this. Okay. Finish your point. And then I want to make a

01:45:10.600 broader point about oncology. So my point is a simple one. The nature abhors a vacuum. So if

01:45:16.580 you say I'm not going to kick off people, the lab is a good example, but I'm not going to remain

01:45:21.760 silent. Here's what I know about that thing, about the lab. Here's the facts, here's the publication.

01:45:28.460 You know, I think that the absence seeds that space to people. I feel like the anti-vax,

01:45:35.200 specifically things like autism, many people have seeded that space on social media because you are 0.99

01:45:41.760 kicked in the butt if you don't. So I do think you can't leave a vacuum. 0.99

01:45:46.360 Yeah. I think that's a great point. I'm glad you brought the ivermectin and cancer thing up.

01:45:50.480 So a couple of my patients, which is a statement, I'm going to acknowledge that my patients are educated

01:45:57.500 and affluent people for the most part. A couple of my patients have sent me clips of various people

01:46:03.580 claiming that ivermectin is curing people with stage four cancer. Now, because they're sending

01:46:09.560 these to me in text and I'm really, really busy, I'm responding in a rather glib way, which is 0.99

01:46:16.200 usually using phrases like, this is effing bullshit. But I usually follow it up a few minutes 0.97

01:46:22.840 later with a text that says, happy to discuss. And usually they say, no, Peter, I just needed to know

01:46:28.180 that this was nonsense. But I also agree that I don't think people should just be taking thing on

01:46:31.940 faith and I really want to be able to offer. So I think I made a note that actually I wanted to

01:46:35.520 discuss this exact example and hopefully we'll be able to clip this particular segment so people 0.99

01:46:39.720 understand why this is such a idiotic statement. To believe that ivermectin cures cancer and to 0.99

01:46:46.940 listen to the stories of multiple people with all sorts of different metastatic cancers that are

01:46:52.160 cured, you're almost explaining that cancer is a single disease. So explain why at face value,

01:46:59.880 the idea that anything could cure multiple forms of cancer is an impossibility.

01:47:07.100 It is an impossibility. There's no doubt about it. Every cancer has very specific biologies

01:47:14.640 that allow it to grow and spread and cause humans problems. And that's why you don't go to the cancer

01:47:22.500 doctor. You go for a prostate cancer, you go for a gastric cancer, you go because the biology of each

01:47:29.420 the cancers is different. And when you go even one step further, as you've alluded to, it's not just

01:47:35.540 that colon cancer and breast cancer are as different as kidney disease and heart disease. It's that breast

01:47:42.120 cancer with an estrogen receptor that lights up versus a HER2-neu receptor that lights up versus no

01:47:48.820 receptors that lights up, those pretty much have nothing in common other than the fact that they

01:47:54.640 originated from the mammary cell of a woman's breast.

01:47:57.340 Right. So we use anatomy to describe where the tumors are. But it is not irrational to use different

01:48:05.360 doses of medicines in combination with other doses. The thing that we went over is the preclinical phase

01:48:13.080 one, phase two, phase three is meant to give whatever cancer patient, whether let's in this case,

01:48:19.760 I think prostate cancer has been talked about a lot, all the information they deserve on both safety

01:48:26.120 and efficacy. Does it work? Does it shrink the tumor? Does it help them live longer? I haven't read

01:48:31.560 anything about ivermectin doing that in patients and what the side effects are and how it could harm

01:48:37.780 patients. So I think patients deserve that kind of information.

01:48:41.280 The other issue I have with this type of rhetoric is the next line that follows is,

01:48:47.940 the pharma companies all know this works. And the reason they're keeping it from you is so that

01:48:53.980 they can make more money giving ineffective drugs. Now, again, I'm going to offer my point of view on

01:49:00.680 this, but you being the veteran of some of the biggest pharma companies in the world, feel free

01:49:04.920 to correct me. I think pharma would be happy to have a drug like ivermectin that cured all cancer

01:49:11.180 because the first thing they would do is put a slightly different modification to it to make

01:49:17.860 it more efficacious, basically less side effects. And they would patent it and they would make all

01:49:24.380 the money in the world. They'd be all over this. If they're able to make a hundred thousand dollars

01:49:28.220 on a drug that extends your life by eight months, I promise you they will be making millions per drug

01:49:33.620 if it's curative. So again, such illogical arguments are put forth and it drives me sort of bananas.

01:49:41.180 But if we want to go back and say, how did we get here? I think when my friend Joe Rogan took

01:49:47.580 ivermectin for COVID, which when Joe asked me, what do I think? I said, Joe, I think it's a totally

01:49:52.920 safe drug. I'm pretty sure it has nothing to do with why you're feeling better today. I think you're

01:49:58.080 feeling better today because you have an amazing immune system. You're an insanely healthy human

01:50:02.200 being. You did 10 other things, two of which might've worked. I'm pretty sure the ivermectin

01:50:07.400 had nothing to do with it. That said, the medical community didn't say that to him. What they said

01:50:11.780 is you're taking horse dewormer, you idiot. Well, that was a strategic error. That's an awful way to 1.00

01:50:17.160 talk to somebody. And ivermectin might be a horse dewormer. It also happens to be, and I look this up,

01:50:23.700 Sue, there is no drug on planet earth that has been taken by more human beings than ivermectin.

01:50:28.140 And it might have the fewest side effects of any drug out there.

01:50:33.080 And look, there may be human conditions for which ivermectin helps.

01:50:36.700 Works beyond, yeah, exactly.

01:50:38.500 Beyond what we know. And I think that's an opportunity for somebody to study it. Good for

01:50:42.540 them.

01:50:43.100 But again, it's something about the elitist nature in which that was handled that has now created

01:50:50.300 this terminal effect of ivermectin is a cause celeb for, I mean, pretty soon someone's going to

01:50:56.760 say it cures Alzheimer's disease, I'm sure. Well, I think it's a drug that's an anti-smarty

01:51:01.700 pants drug. Yeah, that's a great way to put it. That's what it is. To me, that's heartbreaking

01:51:06.000 because the answer should have been, I talked about this with Joe very openly on his podcast.

01:51:10.160 I said, look, I've looked at all the RCTs of ivermectin and COVID. There's no signal,

01:51:15.940 except my memory could be off on this, but there's a little signal in this Brazilian trial,

01:51:20.140 but the methodology of that trial was horrible. So I have to believe this is not working. It's a

01:51:26.260 good try. All about trying. It was a great idea to take off-the-shelf drugs and see if they worked.

01:51:32.600 Nothing wrong with that. We've done that for other things.

01:51:34.460 When they don't, we have to move on. By demonizing it and by demonizing the people that

01:51:38.900 felt it might work, we find ourselves in a situation right now where it's very irrational.

01:51:44.220 One of the things that I've did over the last four years is participate in the President's

01:51:49.340 Council of Advisors on Science and Technology and co-chaired a report on the future of public

01:51:54.680 health. And we ended up thinking that we're focusing on the workforce. One of the remedies

01:52:01.940 for the issue you and I just talked about is having a broader set of people who we think of as the

01:52:09.140 public health workforce. And I think people who are interested in ivermectin, farmers, people who are

01:52:15.020 up close and personal to some of the things with this bird flu, there are a number of different

01:52:19.900 folks who would be really interesting to involve in public health efforts, and we typically don't.

01:52:26.180 And so I think that's one of the ways that we can go forward in public health is to think about

01:52:31.780 how do we define public health and what does it look like?

01:52:34.720 Yeah, I agree with that. Public health has really struggled in some ways. You've had these incredible

01:52:39.520 success stories and then some awful failures. On the surface, it's such a great thing. I think

01:52:46.000 that's why Make America Healthy Again resonates for people. People universally want to be healthy.

01:52:54.180 They want their families to be healthy. This is a universal thing. And how to capture that and make

01:53:00.860 that real, not ivory tower, but real for people who just want their families to have a chance of

01:53:06.960 being healthy. I think that's a real positive. Yeah. I want to talk a little bit about AI.

01:53:11.400 A lot of people might not realize you're on the board of OpenAI, and you're the only person in

01:53:17.440 medicine on that board. So talk to me a little bit about how that came about. I want to obviously talk

01:53:23.380 about the implications of that, what you're excited about, and what you're afraid of.

01:53:27.400 So I joined the OpenAI board almost a year ago now when they had had in November of 2023 what they

01:53:35.960 call the blip, which is CEO fired, board changed over. And I have been so impressed by the intellect,

01:53:48.400 the commitment, the sense of responsibility of folks at OpenAI. I hope this is maybe a little crazy,

01:53:57.600 but here's what I hope. If I had a top two things for AI, one is in some of the things we've been

01:54:04.140 talking about in product development. I mean, I love product development. I think it is the best

01:54:09.420 job on earth. You get to make new medicines for people who are sick. You go home and tell your mom

01:54:14.120 and dad that they're happy. So what if we could take the tool of AI and make easy the things we can

01:54:22.820 make easy? So you don't use AI to change a clinical trial. I still want to know, does your tumor shrink?

01:54:28.800 Do you feel better? Do you have side effects? But there's a lot of study reports. There's toxicology

01:54:34.700 reports. There are a lot of things that are labor and paperwork that are actually very important to

01:54:41.860 establishing the safety, especially, but also the efficacy of a drug. I think using AI more and more

01:54:48.620 on pieces of the clinical trials process so that if something takes time, it's because it's benefiting

01:54:56.020 a human, not because we just couldn't do it fast enough. So the clinical trials, I think, still has

01:55:01.840 some opportunities for that. Give me a time and money sense in terms of savings. This is a very

01:55:08.240 important question. If you said the entire clinical trials program for a drug is six years, let's just

01:55:14.080 make that up. IND to approval. IND to approval. Okay. I would want to cut it down by two years. And you

01:55:19.940 believe AI can do that right now, or we're on the path to that? I think we could be on a path to that.

01:55:24.140 Now, the challenge of it is going to be, if you say, this example I like to give because it makes

01:55:30.620 sense for people. If I'm changing five-year survival, if this is sort of a mature, established

01:55:36.240 thing, I got to wait five years. I can estimate things and I can work with FDA to make sure if

01:55:42.100 people can benefit. And you could argue with a regulatory change in the FDA, if we said greater

01:55:47.180 emphasis on safety to approval, greater emphasis on post-market surveillance for efficacy, we shift

01:55:54.600 this thing a little bit. Now you could say at three years, we're trending, you get a provisional

01:55:59.240 approval, and now we're going to follow you. There's an example, like Pax Levit, in my mind,

01:56:04.380 you could argue maybe should have been pulled. Maybe it wasn't as effective as it looked in the

01:56:08.360 trials. And that doesn't mean they were wrong to approve it because it was any port in a storm.

01:56:13.100 But after the fact, we could have been, oh, you know what? No harm, no foul. It was safe.

01:56:17.660 You can always do that.

01:56:18.640 And so maybe we do that for oncology.

01:56:20.680 I think that the other thing is, you and I both know, if you have 500 patients in a trial and you

01:56:26.760 look at safety, that's so limited. If you have a much more AI-driven, why don't we follow safety in

01:56:32.660 every patient on the dark?

01:56:33.760 Exactly. Ongoing.

01:56:34.960 Ongoing. So I think the opportunities in clinical trials are massive. The other thing I would love to see

01:56:41.080 is a change in the things that cause burnout of nurses and physicians and others in the hospital. 0.92

01:56:47.700 This is across the board, not just in clinical trials.

01:56:49.740 Not in clinical trials. This is healthcare.

01:56:51.620 Yeah.

01:56:52.100 Healthcare should have tools where it's easier to decrease the load, the burden on both caregivers

01:57:01.180 and families. I think that should be doable. It's not that hard.

01:57:05.280 I think that that is absolutely correct. On the nursing front, there's a huge demand, obviously.

01:57:12.520 How much of this do you think of absent robotics? So robots can really change the game. I'm not

01:57:18.260 close enough to that. Are you?

01:57:19.940 I'm not close enough to the robotics piece of it.

01:57:21.980 Yeah. So I don't know how long until a robot is doing what a nurse is doing. But when you think of

01:57:27.940 medical and chart reconciliation and things like that, is that where you think the greatest

01:57:32.860 opportunity is?

01:57:33.840 I think it is when you're trying to connect all the dots. That's the thing. What AI does so

01:57:39.580 brilliantly is it just takes a lot of data and it comes out with observations. And if there are ways

01:57:45.380 that that can assist at the bedside, that's a massive improvement, especially when people are

01:57:51.480 changing, even me, University of Washington to UCSF. It's so hard to change caregivers, to change

01:57:58.300 health systems. Those kinds of things can decrease workloads. But I also think it's the kinds of

01:58:05.000 things where clinical observations could be AI-driven.

01:58:09.860 So the Nobel Prize last year was awarded for protein folding, AI-driven analysis. Explain to people

01:58:16.420 why that is significant. How much do you think that particular achievement is going to advance

01:58:21.980 biotechnology and what remains ahead of it as far as even greater molecule selection?

01:58:27.740 These guys, what they did is they made possible, and we talked about preclinical. This is pre-preclinical.

01:58:35.900 This is how you even-

01:58:36.860 This is figuring out what you're going to do.

01:58:38.100 This is figuring out what you're going to do. If you can make figuring out what you're going to do

01:58:42.540 much, much faster, which they did, you're going to have the opportunity. The way I think of it is

01:58:48.800 you've got like a mountain of opportunity, but it's shown a light on just a limited number of

01:58:55.680 things where you can see the opportunity and take advantage of the opportunity. I think it's a start,

01:59:00.720 but I think it's great that they were recognized.

01:59:03.980 Do you think this is the most important thing from a promise perspective that AI has brought to medicine

01:59:08.600 since-

01:59:09.320 So far. Yeah, I do. So far.

01:59:11.340 And so what do you think would be the next mega unlock? Would it be on the data front?

01:59:16.420 Would it be a predictive model? How could we shorten a clinical trial by 60%?

01:59:23.680 Anything where AI can help us with outcome measures. I told you that my husband's an HIV doc.

01:59:31.340 When we were both at Bristol-Myers Squibb, I was doing two-by-two measurements of tumors on x-rays

01:59:37.180 for Taxol, and he was looking at viral load. Viral load allowed us to have 20 HIV drugs in like five years.

01:59:45.940 It was crazy how good it was. I want a viral load for everything.

01:59:50.900 You need a good biomarker.

01:59:52.520 We need a good biomarker for more things. And you were talking about all the different types of breast cancer.

01:59:59.020 So think about what you just talked about with breast cancer that you have ER positive, ER negative,

02:00:03.900 HER2 positive, triple negative. There's all these. What if actually there's 15?

02:00:08.800 Yeah. There undoubtedly are. We know there are.

02:00:11.800 There probably are. So then you're in 15 trials, but you only need 10 patients in each trial because it's so obvious

02:00:18.040 you have the perfect remedy for each of those patients. I always think of it as switches on, turn it off,

02:00:23.760 and you see clinical benefit. Anything we do that sets up like that, especially if we can not just measure switch on,

02:00:31.400 but switch off. That's why viral load is so powerful.

02:00:34.120 So what's your level of optimism or pessimism around liquid biopsies? And do you think that AI can help us with these?

02:00:43.340 I have been pretty negative based on the data. I just have not seen the data that suggests to me that we're helping.

02:00:51.960 And is this on the sensitivity front?

02:00:53.640 Yeah.

02:00:53.920 Yeah.

02:00:54.260 So can AI help us? Possibly. The problem is just really hard.

02:00:59.020 Yeah. I was about to say, do you think the problem is tumors don't shed enough DNA?

02:01:04.140 I think that appears to be the problem because if they did, I think it would work.

02:01:08.400 So that's the most important problem. The other problem is something that I think we all tend to

02:01:14.080 underestimate because I love the concept of prevention. And I think Make America Healthy Again

02:01:20.400 in part is we'll go to preventive therapy and stop all these. And I understand that in oncology,

02:01:25.640 we've often celebrated tiny successes, but you can't have big successes before you have tiny

02:01:30.420 successes. I don't think it's easy to do early detection. The only two things that are, well,

02:01:38.840 now three, colonoscopy works. For cervical cancer, a pap smear works. Even better, HPV vaccine is my ad.

02:01:47.460 And now you can do a spiral CT for lung cancer. I'm not even using one handful of fingers. And we've

02:01:53.100 been trying to do early detection as long as I've been an oncologist.

02:01:57.760 I agree with you, by the way. I would add to that PSA in the hands of someone who understands what

02:02:02.500 to do with it. So PSA by itself, pretty bad. PSA density, when you know prostate volume and PSA

02:02:08.620 velocity, when you have serial measurements, starts to become very predictive. So you take a man who has

02:02:13.780 not had a prostate biopsy and you stratify his PSA according to PSA density, the ability to predict if

02:02:20.980 he has a Gleason 3 plus 3 or 3 plus 4 or 4 plus 3 is really quite high. And at least you can then

02:02:26.700 stratify those patients more quickly into a PHI or a 4K and ultimately decide, do they need a

02:02:31.980 multi-parametric MRI? And you go down that path. So it's not turnkey. And I completely understand why

02:02:37.340 they've said, we're going to make no recommendation. I do take comfort in knowing. It's sad to me,

02:02:43.140 but I take comfort in knowing too many men are dying of prostate cancer. It should not be the

02:02:48.380 third leading cause of cancer death. And yet I understand that it's a big ask to get every

02:02:54.340 doctor fully up to speed on the algorithm. You know what you just said? That's something that

02:02:59.740 if someone wanted to start a company, they could simplify that and make something more turnkey

02:03:04.560 for patients and physicians. When you go through the four leading causes of cancer death,

02:03:08.600 two of them don't need to be on the list. Colon cancer and prostate cancer don't need to be on

02:03:13.440 the list. They shouldn't be on the list. Now lung, I think we can reduce it a lot,

02:03:16.720 but it's going to be awfully tough. It's tough. And breast is still really tough because it's not

02:03:20.160 Halsteadian. It doesn't have that straightforward progression from polyp to cancer. No, it's true.

02:03:26.420 That's the neat thing is you can just take out the polyp. That's always been the beauty of

02:03:29.860 colonoscopy. What is your level of optimism that we could ever... So instead of just talking about a

02:03:35.560 broad liquid biopsy, let's just talk about breast cancer, what do you think it would take? And do

02:03:40.920 you think it would be a protein? Do you think it would be DNA? Do you think it would be RNA? If you

02:03:44.520 had to guess what would be the earliest signature in the blood of different breast cancers, where

02:03:51.940 would you put your money? I think it'd be interesting to look at protein.

02:03:55.340 Think about how that would change breast cancer treatment. It would. It would be tremendous.

02:03:59.740 It's funny. I say the following deliberately not acknowledging your gender because I'm sure you hear all the 1.00

02:04:04.920 time. Sue, you are the most remarkable example of a woman in medicine. Gender aside, you are just a 0.93

02:04:10.640 remarkable inspiration, period, as a physician, as a business leader, as a public health official.

02:04:16.320 I have been a fan of yours for so long. When I walked into that room last year and saw you sitting

02:04:22.980 there, I was giddy. So thank you for humoring me and making the trip.

02:04:27.080 It was fun. I'm delighted to talk with you. I really enjoyed it.

02:04:30.680 Thank you for listening to this week's episode of The Drive. Head over to

02:04:35.780 peteratiamd.com forward slash show notes if you want to dig deeper into this episode. You can also

02:04:43.460 find me on YouTube, Instagram, and Twitter, all with the handle peteratiamd. You can also leave us

02:04:49.460 review on Apple Podcasts or whatever podcast player you use. This podcast is for general informational

02:04:56.020 purposes only and does not constitute the practice of medicine, nursing, or other professional

02:05:00.640 healthcare services, including the giving of medical advice. No doctor-patient relationship

02:05:05.540 is formed. The use of this information and the materials linked to this podcast is at the user's

02:05:11.720 own risk. The content on this podcast is not intended to be a substitute for professional medical

02:05:16.600 advice, diagnosis, or treatment. Users should not disregard or delay in obtaining medical advice

02:05:22.320 from any medical condition they have, and they should seek the assistance of their healthcare

02:05:26.840 professionals for any such conditions. Finally, I take all conflicts of interest very seriously.

02:05:32.840 For all of my disclosures and the companies I invest in or advise, please visit peteratiamd.com

02:05:39.300 forward slash about where I keep an up-to-date and active list of all disclosures.

The Peter Attia Drive - April 28, 2025

#346 - Scaling biotech and improving global health： lessons from an extraordinary career in medicine ｜ Susan Desmond-Hellmann, M.D., M.P.H.

Episode Stats

Summary

Transcript