Ash Alizadeh, a Stanford hematologist-oncologist and physician-scientist, joins The Future of Medicine for a conversation about cancer blood tests, DNA, early detection, and the growing ability to find traces of cancer that traditional scans may not yet reveal.
Alizadeh has helped pioneer approaches that search for tiny amounts of tumor DNA circulating in the bloodstream. In this episode, he explains the science in simple terms: as cells throughout the body die, fragments of their DNA enter the blood. Cancer cells can do the same. The challenge is finding those rare cancer signals amid an enormous amount of normal DNA — what Alizadeh describes as essentially a “needle in a haystack” problem.
The conversation explores two major possibilities for these technologies. One is detecting cancer in people who have not yet been diagnosed. Alizadeh discusses the promise of multi-cancer early detection tests, but also why showing that a test finds cancer sooner is different from proving that screening ultimately helps people live longer. He argues that these approaches may prove particularly useful when focused on people already at higher risk rather than trying to use a single test equally across dozens of cancers.
The second application may be closer to transforming cancer care: looking for molecular residual disease after treatment. Once doctors know the specific DNA changes within a person's tumor, a blood test can search for those same changes later, potentially revealing that cancer remains or is returning before it becomes visible on a scan. Alizadeh discusses how personalized DNA tests are already being studied and used across cancers including bladder, colon, lung, and breast cancer — and how emerging evidence could help doctors decide who needs additional treatment and who may be able to avoid unnecessary therapy.
Together, Alizadeh and host Euan Ashley explore what cancer DNA can tell us, where early detection still falls short, why finding residual disease could change treatment decisions, and what it may take to make cancer monitoring far more precise.
The Intro
“She showed up without her hair. She'd been diagnosed with ovarian cancer in Iran.”
Dr. Alizadeh is a cancer genomics pioneer at Stanford. When he was a teenager, his family left Iran to escape the war. His mother followed months later and when she arrived, she was diagnosed with cancer.
“For most cancer, the proportion is much much less than a percent. Yeah. So really a needle in a haystack problem.”
It became the basis of his life's work and led to the creation of a new kind of blood test, a liquid biopsy that can find a cancer's DNA in a few drops of blood, catching recurrence long before it shows up on a scan or even detecting a range of cancers early.
“The mutations are so private that they tend to be a great barcode to keep track of disease.”
In this conversation, we trace how a kid from Tehran became one of the architects of the genomics revolution and where cancer detection is headed next.
Am I going to be able to see my kids' graduation? Am I going to be able to attend the wedding?
Welcome to Stanford Department of Medicine's inside look at the future of medicine.
The Transcript
Dr. Euan Ashley: Well, Ash, thank you for joining me here on The Future of Medicine.
Dr. Ash Alizadeh: Thank you, Euan.
Dr. Ashley: We were very happy to have you join us for our Medical Grand Rounds and you gave a really stunning presentation taking us through a lot of the work that you have pioneered, and really rarely do we have a guest who has essentially, with some help, but essentially built a field.
Dr. Alizadeh: I don't know about that but thank you very much, standing on the shoulders of giants here.
Dr. Ashley: Well, we love to talk about some of those giants, but also talk about the work that you've been doing. But before we get to that, we always love when we have guests on to talk to them about their background. What was it motivated you to come into medicine? How did you end up at Stanford University pioneering the future of early and recurrent diagnosis of cancer using cell-free DNA? So, let's go back to the beginning.
Dr. Alizadeh: Oh, sure. So I came to Stanford in 1994 to start medical school here kind of undecided about whether I would do medicine for medicine or a physician scientist type path.
Dr. Ashley: But take us back further, right? When did you first want to be a doctor?
Dr. Alizadeh: Oh, that's a good question. Probably the experience of, as a teenager, my mom's cancer was probably the thing.
Dr. Ashley: Yeah, tell us about that one.
Dr. Alizadeh: So, my family immigrated to the US when I was a teenager during the Iran war. And I had to leave just before my 14th birthday because that was the age at which you could be conscripted. And so we left just a couple months before I was turning 14 and came with my dad. And when it took longer for my mom to come than we'd expected, I knew something was off, but I didn't know what. And then she showed up without her hair. She'd been diagnosed with ovarian cancer in Iran. And I remember her coming here and kind of hearing about, you know, what ovarian cancer is and that she hadn't had sufficient therapy and that she was likely going to die of it and kind of being in the dark. It was even before the days of CA125 biomarker testing. It was just beginning around that time.
Dr. Ashley: Yeah.
Dr. Alizadeh: So it was kind of a sense of doom and gloom, but it got me interested in science I think to a certain degree.
Dr. Ashley: Yeah.
Dr. Alizadeh: Though I'm not 100% sure how, but attended college at UCLA, studied biochemistry there, and I started in a lab from very early on, genome technologies basically. I worked for a scientist there named Winston Salser who had his hands in the founding of Amgen.
Dr. Ashley: Right.
Dr. Alizadeh: And then came to Stanford.
Dr. Ashley: Yeah, that remarkable time. Where did you live when you first moved to the States?
Dr. Alizadeh: Orange County, actually, Southern California. But where a lot of Iranians congregate, right?
Dr. Ashley: I have heard that.
Dr. Alizadeh: Right. Back when I was there, I think there was only one other Iranian kid with me in school. So it was kind of interesting.
Dr. Ashley: Pretty big culture shock for a teenager.
Dr. Alizadeh: Exactly. Exactly.
Dr. Ashley: Were you from Tehran?
Dr. Alizadeh: Yes, my family. Big city. Yeah. Grew up in Tehran.
Dr. Ashley: Right. So at least the big city part was familiar.
Dr. Alizadeh: Yes.
Dr. Ashley: Coming to near LA.
Dr. Alizadeh: Yes.
Dr. Ashley: But otherwise not too similar.
Dr. Alizadeh: Not too similar.
Dr. Ashley: Yeah. So then UCLA, then you ended up at Stanford.
Dr. Alizadeh: I ended up at Stanford medical school.
Dr. Ashley: Yeah.
Dr. Alizadeh: For medical school, and I wasn't totally sure if I would want to do the MD track or the MD PhD track. Yeah. But remember hearing some pretty amazing talks as a first year student. Heard Ron Levy give a talk for medical students, Irv Weissman give a talk, and I think a talk that really caught my attention was Pat Brown, right? And then I had a genetics class with David Botstein. This was kind of the, yeah, dawn of the genome era but for model organisms, before the genome project was complete for humans.
Dr. Ashley: Yeah. And Pat Brown of course, famously lots of things, but invented the microarray, one of them.
Dr. Alizadeh: Yes. So yeah, he had just invented the microarray which is a dated technology now. But I remember having done a lot of northern blots as a student at UCLA. Seeing that they had studied the cress, Arabidopsis thaliana, to study the tip of the root and the tip of the shoot, comparing these two parts of the anatomy of this plant. Yeah. I think it was like 100 or 200 genes at once. That's magic. And so I wanted to do that, to be able to study all genes at once. So I approached Pat and said, "I want to study human beings." And they said, "Well, we don't even know how many genes there are."
Dr. Ashley: Yeah. They thought there was 40,000 or 70,000.
Dr. Alizadeh: I went to the Cold Spring Harbor, there was a genome meeting that we used to all attend. There was a bet I remember and I bet at 140,000 and I think the average was about 100,000.
Dr. Ashley: Yeah. I think I still have a genetics textbook that estimated 100,000.
Dr. Alizadeh: 100,000. But at the end it wasn't so many, but Pat was gracious enough to take me under his wing. I spent a couple years at the NIH and then came and finished my PhD at Stanford with Pat Brown.
Dr. Ashley: Yeah. When would that have been?
Dr. Alizadeh: I started in 96 and finished in 2001, my PhD.
Dr. Ashley: Yeah. Right. Yeah. 2002 was when I first arrived, and at the time I remember Pat's microarray was there. We had been working in our lab here with Agilent. Yes. And had a microarray that I think had as many as 3,000 spots on it, which you know, given the earlier ones had been a few hundred, but that's like a few hundred more than one.
Dr. Alizadeh: Yes. Yes.
Dr. Ashley: So it was an amazing time. And then over those few years, they went from 3,000 to most of the genome, being able to cover each gene. Very exciting time in technology as it continues to be.
Dr. Alizadeh: Right.
Dr. Ashley: Yeah. So that gives us a sense of how you came to medicine, how you came to genetics. Obviously some of these, David Botstein, Pat Brown, these are very big names, very inspiring individuals. Irv Weissman you mentioned as well. Yes. And so you did residency.
Dr. Alizadeh: Yes. I joined medicine to do my medicine residency and then hematology and oncology fellowship here. Had some really great teachers. Yeah. And ended up doing some postdoc work with Ron Levy and Irv Weissman. Got some things done with them. But yeah, learned a lot from all of them.
Dr. Ashley: When did you first hear about cell-free DNA, and maybe for our listeners who are not familiar, because we have a broad range of listeners, explain what that is?
Dr. Alizadeh: Yeah, so I had never heard of it honestly before, as a student or even in my residency, to really know what it is. I think we had done a journal club when I was a resident, and Gilbert Chu, one of our biochemistry and medicine oncology faculty, was talking about a lung cancer study where they were studying DNA in blood, and I thought this can't be real. What are you talking about? But it was in New England Journal of Medicine and there was a supplemental figure. And then it was during my, yeah, was it? Yeah. It was first year of fellowship. Yeah. With my second child, we needed an amniocentesis and I remember thinking, whoa, this is brutal. Yeah. And you could lose the pregnancy, and the reason we're doing this is because we've had some indirect measures of a genetic abnormality for Down syndrome. Yeah. How thick the neck is, the stripe on the back, and then whether some proteins are abnormal. But now we need the real DNA for an extra copy of a chromosome 21 and we need to get in there with a needle and get it. And so I thought that was interesting but I didn't know how much could be, and by chance I attended a seminar that Steve Quake was giving, right, and I thought, well –
Dr. Ashley: Recent guest on our podcast here.
Dr. Alizadeh: And I thought this is amazing, but it was still not widely available, it was just starting to be. And I thought okay, the idea there in prenatal diagnostics, for those who didn't see Steve's talk, is this revolutionized this field basically, of instead of doing these indirect measures of the genetics using proteins and ultrasounds and things like that. That now hundreds of thousands or millions of women each year get these tests as better tests, for a blood test.
Dr. Ashley: A blood test. Yeah. The baby's DNA floating around in the mother's bloodstream.
Dr. Alizadeh: Exactly. And it's astounding actually when you think about it, because by a few weeks of pregnancy something on the order of 5 to 10% of the DNA in mom's plasma is from the fetus. Not because 5 to 10% of the cells circulate in the blood, but because that fetal tissue, the placenta and the other organs that are turning over, are just spewing out that DNA. And so it's like shooting fish in a barrel, so to speak. Yeah. To detect an extra copy of a chromosome. And Steve invented the method for doing that that's commonly used today.
Dr. Ashley: Yeah. And we talked during that episode about the remarkable revolution from this invasive test with literally a mortality rate. Yes. To the point where millions of women around the world are now having this non-invasive test. And in those early studies, and I think he often quotes a study from a long time ago, almost 100 years ago I think, when the idea of nucleotides in the bloodstream first came about, that speculated at the time that perhaps this may be a method of monitoring both pregnancies, yes, and cancers.
Dr. Alizadeh: Yes. Yeah, two French scientists actually, I think the main one Mandel, in the 1940s in occupied France, as a Jewish revolutionary, kind of a Jewish freedom fighter that left Poland I'm pretty sure and was actually part of the partisans, amazing, ended up doing some science on the side and briefly after the war published the paper that Steve showed describing free nucleic acids. Yeah. But this is before it was even completely clear that DNA is truly the genetic material, the structure of DNA hadn't been solved, right, and so kind of amazing.
Dr. Ashley: But it's very early.
Dr. Alizadeh: Very early. And a lot of the tools we have now we take for granted honestly, like the ability to make copies of DNA, to sequence DNA, to do all the things.
Dr. Ashley: They had none of that.
Dr. Alizadeh: Right. So they had some very basic biochemical methods to say this is nucleic acid, but they made the observation that there's free stuff in the acellular plasma or serum in the blood, and that kind of gives you a window into the other cell types that don't circulate, the tissues that we are interested in.
Dr. Ashley: Yeah. Because I think that most people when they think of blood think about red blood cells or white blood cells, and then they know that there's plasma and people have given blood and seen that, but you know, and then I guess at some level we know there are proteins that circulate that could be measured, but there hasn't really been much attention paid, despite this very early observation, to the potential diagnostic utility of the circulating DNA and even, as Steve mentioned, RNA. Mhm. But that revolution in the ability to pick up potentially cancer early, and certainly particularly at the moment recurrent cancer, and we'll talk about that, is one that I think really its time has come, and you've really been one of the leaders in the field. So draw us a line from seeing a presentation by Steve and having to undergo with your wife an amniocentesis to turning your lab work, maybe initially as a postdoc and then eventually your own lab, to this area.
Dr. Alizadeh: Yeah. So I think it was mostly incubating an idea. Yeah. And honestly my job talk didn't have circulating DNA. It did not. Too risky. Yeah, there was some early work from luminary scientists, Vogelstein and some others, that had done some work in this area, but it was still pretty early days. And we were trying to apply some of the ideas from the prenatal world, from whole genome shotgun sequencing, to kind of focus on the business end of the cancer genome by focusing on genes of interest. So that was the idea, like what we might try to do. But one of the people I did my PhD work with in the same lab, Max Diehn, who was a radiation oncologist focused on lung cancer, I was just finishing my training, oncology training, taking care of lymphoma, we both chatted with each other about the idea of joining forces to tackle this problem together. And it became a great partnership. So seeing the patients we were seeing, consenting them to go on clinical studies to learn from their blood as we're treating them, so that we're learning while we're stumbling in the dark essentially.
Dr. Ashley: Right. Right. So is it, again for our listeners who are not familiar, is it hard to measure DNA in the bloodstream? What special, yeah, methods do you need to do that?
Dr. Alizadeh: Yes. So just for big picture, in a milliliter of blood you have a few million cells worth of DNA in the cellular portion of the hematopoietic blood cell types. Yeah. And in the acellular portion of blood you have a few thousand cells worth of DNA and the other parts. So several orders of magnitude less DNA. Yeah. But that DNA is the experience of dying cells from all over the body. Still mostly hematopoietic, about 60 to 90% hematopoietic. But the portion that's not hematopoietic gives you a window into those other tissues. So you can measure it. It's in the order of nanograms of DNA instead of micrograms of DNA per milliliter. And then a proportion of that DNA is cancer derived. If you're thinking about cancer, for most cancer, the proportion is much much less than a percent. Yeah. So when we were talking about the fetal application, I mentioned fish in a barrel because it's in the high single digit, low double digit when you're trying to call an extra copy of chromosome 21. We're dealing at much lower partial concentration. So really a needle in a haystack problem.
Dr. Ashley: Yeah. Right. So technically there were some hurdles to overcome to just be able to find this signal from the noise. So talk a little bit about that, and then tell us how you fired the, so yeah, I remember the weapon when you built it.
Dr. Alizadeh: Well, when we first started I think there was this idea that you needed micrograms of DNA to make a library. So I thought that this is crazy, that cannot be, and we knew there were companies working on doing single cell libraries for RNA. So if that can be done, we are going to make a good library from low input material and have it be a faithful library. So that was kind of a molecular biology and kind of the Pat Brown style, right? I just want this done kind of thing, and it's going to get done.
Dr. Ashley: Yeah. Just keep iterating.
Dr. Alizadeh: Iterating. And so that was through some really talented people that worked very hard, Scott Bratman who's now at Toronto, and another very talented fellow Aaron Newman who's in our biomedical data sciences, working as a team together to kind of solve these problems, to make really good libraries and then to pick the business end of the genome. We focused on lung cancer as our first example, and picked a very small sliver of the genome to focus on, and to do targeted panel sequencing on very small libraries. And that was an advance, and it's really common now. Everybody does the same method, but that method was called CAPP-Seq. It's analogous to all the methods that are now used for blood-based genetic non-invasive genotyping of tumors of interest, and you can do that from the blood in the advanced setting, but you can also track mutations if you know about them in the tumor and the blood for monitoring a tumor of interest. The mutations are so private that they tend to be a great barcode to keep track of disease.
Dr. Ashley: Yeah. Well, that's, you know, something that we'll come back to, I think, but maybe a good moment to explain that a little bit, because for you to be able to detect cancer in the bloodstream, you need a signature. Yes. Something that's different from regular DNA.
Dr. Alizadeh: Yes. So, I like to use the analogy of Waldo. I think most people know the Where's Waldo analogy. If you know what Waldo looks like, you know, the spectacles and the hat and the striped outfit, that helps you find Waldo in the book. That knowledge is analogous to mutations that you know from a tumor, and that's what we call tumor informed tracking strategy.
Dr. Ashley: So you've sequenced someone's tumor, you've found an abnormality by comparing it to their normal.
Dr. Alizadeh: That's right.
Dr. Ashley: And then you can now track the tumor DNA in the blood.
Dr. Alizadeh: That's right. And with that kind of knowledge, having a tumor informed tracking strategy, especially because you not only know about Waldo's glasses, you know about the hat and whatever, that combination, and then you can take mutations by that same analogy, keep track of them, you have a lot of sensitivity. You need one Waldo in a very big crowd that you can find. That's tumor informed. The corollary is a tumor naive, or you know, if you don't know that somebody has cancer and you want to detect cancer early you don't have that luxury, right, and it's a tumor naive strategy. If there were tons of Waldos and they stand out from the rest of the crowd, without you knowing what Waldo looks like you could still tell that Waldo is relatively strange, but you generally need higher concentrations or a larger number of that type of species than if you have some prior idea of what they may look like. So those are the two things that we coined back then, and I think they've kind of stuck.
Dr. Ashley: They really have. I mean they've become big businesses as well, although I think with quite differing levels of success depending on how you define that at the moment. But I think these two strategies also map to measurement of recurrence of cancer, because in order to understand what a tumor informed test would be, you have to have diagnosed it to begin with. So that's in a patient who'd already had cancer, and you can map their therapy and any recurrence. But there's also a lot of interest, as you just talked about, in the early detection of cancer among otherwise healthy people, and of course everyone wants to detect cancer early, and the idea of a molecular test for cancer rather than an imaging test or a test for blood in the stool for example is very appealing. So lots of investment, biotechnology investment, large companies, small companies, in this idea of multi-cancer early detection. Now you did touch on this in your talk, but give us a sense of how that differs. You already sort of began a little with the Waldo analogy, but how do multi-cancer early detection tests work, and how well do they work today?
Dr. Alizadeh: Yeah. So I have to be careful a little bit here, because some of my dear friends started companies around this. So I haven't spent as much effort on it as they have, but I kind of know of the literature. The way I think of these tests is they still by and large look for DNA, and the most commonly used multi-cancer detection tests look for not DNA mutations but modifications to DNA, a change called DNA methylation, CpG methylation, an epigenetic change to DNA that tends to be very characteristic of cancers and tissues, and when you know what that typically looks like, it's not entirely tumor informed, but it's templated by what the average lung cancer looks like, what the average pancreas cancer looks like.
Dr. Ashley: And these are chemical changes to the DNA.
Dr. Alizadeh: Chemical changes.
Dr. Ashley: Across the genome.
Dr. Alizadeh: Across the genome. And so when you take a healthy subject's blood sample and you look for these changes in the sequencing of DNA for the portion of the genome you're interested in, you don't find so many of them. But in patients who have cancers, you find signatures of the tissue of interest, by and large a tumor of interest, and the characteristic modifications. So there are a few of these that are commercially available.
Dr. Ashley: And just because probably our listeners will have heard of some of them. Grail has one.
Dr. Alizadeh: Grail has one, and Exact Sciences has another one. These are both methylation based primarily, multi-cancer detection tests. That Grail one is called Galleri, and I think the Exact one is called Cancerguard, the commercial name. Yeah. And they've been applied to kind of being a Swiss Army knife for detecting many cancers.
Dr. Ashley: How many cancers do they detect?
Dr. Alizadeh: I don't remember the Exact one, but the Grail Galleri one, 50 cancers is the specification, so up to 50. And then there's a subset of the cancers that are called high signal cancers, which tend to be easier to detect, and some of them are pretty aggressive cancers, that are disproportionately detectable among the 50. And so the biggest studies have applied this in populations at average risk. So we haven't seen the publication, but for example the Grail test applied in a big British trial called the NHS-Galleri study was trying to see if you could detect cancers earlier, to move stage three and four cancers into one and two as the primary endpoint across all those cancers.
Dr. Ashley: So at the time they were detected, they would be detected earlier. That's when you can do more.
Dr. Alizadeh: That's right. Now that is, you know, some might say that's a great initial move, that you have to detect cancer at earlier stages if you're going to improve the outcomes, but it doesn't prove that treating an earlier stage cancer that's detectable is likely to improve outcomes. So some have very vociferously advocated for other harder endpoints like survival or other things, just to say you run a study where one group gets the test, the other does not.
Dr. Ashley: You follow them over time and you literally measure how many live or die.
Dr. Alizadeh: That's right. And there are good reasons for that. So for example in ovarian cancer, the cancer my mom survived, there was a big British trial of a quarter million women screened for epithelial ovarian cancers with a couple protein biomarkers, and of course this test, or this combined strategy, was able to detect ovarian cancers earlier but unfortunately did not improve survival. So that was called the UKCTOCS study. So that was the concern, that this study might move stage but not improve outcome, but the initial attempt was let's walk before we run, and in that attempt unfortunately the primary endpoint that was recently announced was not met. So for the broad group of all cancers they could not detect them earlier, but for the high signal cancers a secondary analysis suggested they might, but I think that still needs more work.
Dr. Ashley: Yeah. So it seemed like, and this was variously reported and obviously discussed, and it's one of the things that is intrinsic to the way we design clinical trials. In order to control the variability and the error, you have to define your endpoint in advance. And so then the study fails or succeeds according to that definition. And there's obviously other analyses that can be done that can show signal, but since you've kind of been forced into making the decision as to what the primary endpoint is, if it doesn't hit that, the study is viewed as a failure, regardless that maybe you chose the wrong endpoint, but there's no way to know that. So what's your own take though on the technology? I mean, we'll come in a moment to talk about MRD testing.
Dr. Alizadeh: Yeah, I think that they show promise, but I think we have not applied them to the place where they might best shine. Yeah. Which I think is to focus on the at risk, higher risk populations. Where, just the same way we would not screen for cancers in children as much as we might in adults, or ovarian cancer in a male or prostate cancer in a female, you could highlight a higher risk based on exposures, based on heritable risk, and then focus to see if in that population you could make an incremental advance. And do you really need a Swiss Army knife for 50 cancers? Could you imagine having a sharper tool, a better knife for one task than the Swiss Army knife for all tasks? That's an open question. We've kind of favored the idea of building a different tool for each application, but it remains to be seen. These are much bigger. The business world has a lot more to say about it than, yeah, than we do.
Dr. Ashley: So right. Now some of the other tests go beyond DNA methylation, still in the early detection world, and there's a sense that maybe that could help provide some additional power, and some data, again I'm not sure if it's published yet, from the Exact scientists.
Dr. Alizadeh: Yeah. So Exact has had an application where they've incorporated proteins at the same time. If you know the Exact main product, the Cologuard test that some of our listeners would have experience with for detecting colon cancer, is a multi-analyte detection test, a protein measurement, a DNA mutation, a DNA methylation, DNA fragmentation. So it's a multi-analyte test. Versions of that have been applied for blood. But admittedly, I have not seen really strong compelling data that the integration of these multiple biomarkers truly is transformative for overcoming what we experience, and I think that I've mostly seen incremental additions from the addition of something else. Are there really orthogonal signals where you cover a base that the other method did not, especially because they become more complex and there's only so much blood you can collect. Yeah. And more expensive to measure four or five things. So proteins, immunologic signatures is something else folks have looked at, DNA mutations and methylation at the same time, RNA and DNA. There are all kinds of ideas there, right? And we've played with these ideas in the lab, but I must say that I have yet to see a knock your socks off kind of demonstration. Yeah. That we've solved a problem that we need to see.
Dr. Ashley: Yeah. I think there clearly was a lot of hope and clearly hype around the Grail Galleri test. I think the story is not finished, for sure, but it was obviously an enormous investment in time and energy to develop the tool and then to run a very large trial in the UK. So I think disappointing for many. But it sounds like your interpretation is still that there is signal to be found, and I think there will be a molecular early detection test that will break through at some point. Yeah. To switch then to an area that seems to be very promising, and an area that you in particular have worked on. You talked about the tumor informed testing, and clearly that's a different proposition when you already know some of the mutations that are in the tumor and you're looking for recurrence. This goes under the broad heading of minimal residual disease or molecular residual disease, or any of the four or five different possibilities for the three letters MRD. But this kind of testing has been quite successful, and your technologies have been incorporated in a number of these testing scenarios. So talk a little bit about that.
Dr. Alizadeh: Sure. So as a hematologist we kind of take these things for granted. So one of the poster children for a hematologist in terms of precision medicine kind of applications is a disease called chronic myelogenous leukemia, where one of the earliest targeted therapies was developed for targeting a tyrosine kinase called ABL, and that drug was very potent at suppressing the disease, and molecular evidence of the disease after the cells disappeared from the blood. The spleen shrank, the bone marrow normalized. You couldn't see the bad cells anymore, but the molecules were left behind. Kind of defined this thing called minimal residual disease, or molecular residual disease. And that paradigm as a way to keep track of response extended to other blood cancers. So in the context of therapies where the tip of the iceberg is below the surface, we no longer can morphologically see the cancer, can't see it on an imaging study, can't see it by flow cytometry, but the more sensitive molecular techniques can find, especially with the polymerase chain reaction, to amplify the molecules of interest and to look for those things. So that idea of applying that to solid tumors and using DNA mutations was something we pursued, and the idea of keeping multiple aspects of Waldo, the glasses, the shirt, the hat, the scarf, and as many mutations to keep track, and just sampling statistics, as you know, you don't need to see all of them but enough of them. So that tumor informed strategy has been used I think most successfully by some of the companies.
Dr. Ashley: And we should mention probably that both of us have actually started companies, exactly, that have tests in this domain.
Dr. Alizadeh: Absolutely.
Dr. Ashley: So clearly we believe in the power of this technology. But yeah, talk about some of the data that we're beginning to see from these technologies.
Dr. Alizadeh: Yeah. So in a range of solid tumors now, there are an inordinate number of tests being done by using tumor informed customized panel sequencing to track disease for bladder cancer, colon cancer, lung cancer and breast cancer. There are now Medicare reimbursed tests for monitoring these personalized panels. Whole exome sequencing is done on a primary tumor. Primary therapy is done. The patient has a resection, but we know we can't always be sure that there isn't residual disease, and having an informed discussion with the patient, the same way as my mom and I had to think about, you know, her survival. You know, can you actually have this be an informed decision on whether you get an adjuvant therapy or additional therapy or other things beyond the surgical management of your disease. So those are amazing tests, and they started with exomes and now with whole genomes, where you know, instead of 2% of the genome you look at 100% of the genome effectively, and then many many mutations to keep track of. These really are amazing tools.
Dr. Ashley: So it's in part a sensitivity question. We talked about how doing a molecular test will give you more sensitivity than say someone going for a CT scan for recurrence of their tumor, or to reassure them that they're still in remission. Now we have the opportunity to look at a molecular test that is personalized to their own tumor.
Dr. Alizadeh: Highly bespoke, a tailored test to keep track of the disease with whole exome or whole genome methods, and looking for the iceberg when the tip is below the surface. And so these tools have very strong prognostic value, and they beat imaging. But a key question has been, do they actually improve outcomes? Is learning bad news early, can you do something about it? And now there's emerging evidence that you can. Yeah. Probably the example that's caught my attention the most is in bladder cancer, for the choice of adjuvant immunotherapy after resection, there's about a year advantage in survival if you do that in a tumor informed MRD setting. So that is surprising, but gives us a sense that as we're walking now, we can run.
Dr. Ashley: Yeah. So this is a decision as to whether you should get extra therapy, in this case chemotherapy, after having cleared a tumor after surgery.
Dr. Alizadeh: Right. That one actually, I think the study there was adjuvant immunotherapy with a checkpoint antibody, in muscle invasive bladder cancer. Yeah. But the same kinds of designs are there for lung cancer, for breast cancer, for lymphomas. We're doing some studies where you've done a definitive therapy, you think you're done, and generally we wait till the patient relapses before we pull the next, right, arrow out of the quiver. Here the idea is, well, in those patients who have residual disease, in which the predictive value is super high for recurrence, 80, 90, sometimes above 90% likelihood of failure, instead of waiting till the disease manifests as a metastasis. Yeah. To actually intervene when the residual disease is small. Yeah. With drugs hopefully that are not toxic but that can eradicate that minimal residual disease.
Dr. Ashley: Yeah. It just makes so much sense, you know, before waiting for a disease to come back and then spread. Yes. That we find it when it's just a small number of cells releasing a small amount of DNA. Yes. It seems like exactly what everyone would want. For me, some of the most striking findings were, for patients in some trials, if the minimal residual disease test, the DNA test, was negative, you know, that was literally 100% predictive of remission.
Dr. Alizadeh: Yeah. That's also for a long period. Yeah, the negative. Yeah, you're absolutely right. And maybe those patients could get less treatment, and you know, maybe they didn't need chemotherapy before their surgery in the neoadjuvant setting, or some other questions. Do we need less intense therapy?
Dr. Ashley: Yeah. So that seems like, you know, hopefully benefits on both sides, and that's what we're all hoping to do in the science work that we do and move forward. We are on a podcast called The Future of Medicine. So, and we've talked about these tests are all available today, you know, even as we build the evidence base for some of them. You've mentioned some of the MRD tests being reimbursable by CMS, by Medicare. What does this look like going forward though? If you take us five years into the future or 10 years, what does the landscape of both early diagnosis and treatment of cancer look like?
Dr. Alizadeh: Yeah, I think, just the same way as prenatal diagnostics sharpened our ability to make informed decisions around risk in pregnancy, I think we're going to be doing the same thing in cancer. It's a harder problem because we have so many different types of cancers and genetic changes and types of therapy, but I think we're going to be moving in that direction. I think we're going to learn how to complement blood tests with imaging so that we're walking less in the dark. Yeah. With the imaging tools, to kind of use them more wisely. And I think we overuse, I think imaging is fantastic and I depend on it as an oncologist every day. Yeah. But we overuse imaging. I don't know if you know, we do some 90 million CAT scans a year in this country. Mhm. That's like one in three to four people getting a scan a year. Not all that is for cancer, but there is an attendant risk with the associated radiation. Some people have argued that a few percent of all cancers are going to be associated with the thing we're using to try to monitor it. Yeah. So maybe we could figure out how to use that imaging in a more judicious way. Yeah. By having it be informed by good tests, the blood tests. But I really think that walking in the dark, I think it's one of the hardest things I've had to do with my patients. Doctors and patients are both reasonably good at dealing with uncertainty, but this type of uncertainty after you've given a certain treatment. Just in reflecting on my own experience with my mom, I kind of see it every time you finish a treatment or a surgeon has done their job with their tumor. The uncertainty is so difficult to deal with. Am I going to be able to see my kids' graduation? Am I going to be able to attend the wedding? What's the probability? And it's not so precise. And I think these tools will help that discussion and the choice of tools we use to monitor and treat any residual risk. I think that will be auspicious.
Dr. Ashley: Well, you brought us back to your mom, but I was going to ask you what had happened with her story.
Dr. Alizadeh: Yeah. So I think I remember visiting a couple of oncologists and thinking, oh wow, she got this medieval treatment in Iran during a war we were fighting at the time, and she didn't finish it, that she would recur, but miraculously she didn't. She had a number of complications of her therapy, because of the types of chemotherapy, but she survived it. Yeah. But the anxiety of when this cancer is going to come back is something we dealt with. But she's done fine. She's done fine.
Dr. Ashley: Oh, that's great. Really great. She must be very proud of the work that you've been doing and moving your field forward. Thank you so much for being part of our faculty here at Stanford Department of Medicine, for the incredible work you're doing for the world, and thanks for joining us on the podcast.
Dr. Alizadeh: Thank you. Thanks so much.