Manju George 0:00
Today, as part of Doc talks, we have Dr Stacey Cohen with us. She’s an associate professor at the University of Washington and an associate professor at the Fred Hutch Cancer Research Center, and she’ll be telling us all about circulating tumor DNA and its role in colorectal cancer. Thank you so much for doing this.

Dr. Stacey Cohen 0:19
Thank you. It’s a wonderful opportunity to get to speak with you, and I’m happy to talk today about circulating tumor DNA, which I’m calling a new frontier in colorectal cancer. Here are my disclosures.

Dr. Stacey Cohen 0:34
So we’re going to go over CtDNA, what is it? How do you detect it? Thinking about it as a marker of molecular residual disease, and then finally, it’s use in active cancer. So first taking a step back and what is CtDNA? So if we go at a really basic, basic level, inside every cell, there’s DNA gets converted to RNA, and then proteins, and from there, we can use those as potential, what we call biomarkers, or ways to try to sample what’s going on- that’s all within the cell. And then, if you take a single cell, and it’s a cancer cell, and it replicates and makes a tumor, then we’ve got a ball of cells there. So if we’re actually trying to sample or biopsy something, we could do our classic tissue biopsy, and then we have,a needle being put in, or a surgical biopsy, where we’re taking a piece of that ball of tumor. Or we could use blood based markers. So this could be something like CEA, which is a protein that’s made by the cells. Or the so called liquid biopsy, looking for circulating tumor DNA. And so this is a newer concept, because there you’re actually looking for the DNA of the cancer cells themselves in the blood. So in everybody, we have cell free DNA that can be detected, even in healthy individuals at low levels. And when it was first done as a screening tool, it was applied to prenatal screening- look for abnormalities in the numbers of chromosomes. So as an example, it could be used in a pregnant woman to look for trisomy 21 which is the hallmark chromosomal change of Down syndrome. But we now apply it in cancer to look for circulating tumor DNA, or CtDNA, which is the idea of looking in the blood for the mutations that we would expect to see in cancer, specifically.

Dr. Stacey Cohen 2:19
The reason why this could be very interesting is it gives us a real time sample of cancer activity. So it has a half life, or how long it’s around in the blood of less than two hours. So it changes very rapidly. And we know that there are some things that are good about it and some things that are limitations. So it is higher at higher levels of cancer. So we disay it’s cancer burden dependent. And there are ways that you could get potential false positives. So when we think about looking at CtDNA, there’s a lot of different ways that we can do it, but because it’s relatively easy to obtain, we can actually check this more than once. So we could do this early on.

Dr. Stacey Cohen 2:59
So this is kind of a timeline diagram here. So we could use it for the detection of cancer in the first place, hoping for screening or earlier diagnosis. We could think about it after a cancer is found, for molecular profiling or prognostication, to have an estimate of how someone might do with cancer. After surgery or definitive management, we could think about for detection of residual disease. If someone’s on longer term therapy, we could use it to monitor response to treatment. And then finally, we could do something looking for what’s called clonal evolution, or looking at the profile of different mutations in the DNA, and see how that changes over time, to try to see if that will help us better understand how to treat a patient.

Dr. Stacey Cohen 3:45
So when we think about methods for detection, there’s various different ways of doing this. There’s our basic PCR based approach, where you pick specific targets and it has fairly good sensitivity. Meaning that you don’t need very much of it to find it. And then we’ve got our fancier technology, next generation sequencing, where compared to the one target of PCR, we might look at over 1000 targets in this situation, and has much better sensitivity, though there is a range depending on the assay. Or you could do an even broader look, what they call whole genome or whole exome, where you look across the whole area, as opposed to specific area., And then you look at it with a little bit less sensitive approach. Within this we can also think about ways of detecting it that’s either what they call tumor naive so you’re checking the same sample or the same assay for every single person. Or tumor informed where you’re actually taking someone’s specific tumor sample and trying to find specific markers from the tumor sample that are also present in the blood. And with the data that we have so far, there seems to be some correlation between an individual’s tumor and finding it in the blood that might be a little bit more sensitive for molecular residual disease detection. Though, as I’ll talk about, as the assays get better, then that might not necessarily be the case.

Dr. Stacey Cohen 5:07
So this is not necessarily information that every patient needs to know, but I put this up here to say that there are pitfalls and limitations of CtDNA, just as there are with everything in medicine, and if something either isn’t collected correctly, or the person that’s ordering the test doesn’t know how to interpret it correctly, we can’t really use the information as readily. So again, important concepts are that it needs to be collected in the right types of tubes. This is not just a regular blood test. You have to know that you’re going in to order it. That the more cancer that there is, then the easier it is to detect it. And if you’re looking for really rare mutations, but you don’t have the type of test that’s looking for that, then you’re not going to find what you’re looking for.

Dr. Stacey Cohen 5:52
So as an example, this graph is supposed to show how different amounts of cancer in the body can actually affect the ability to detect CtDNA. So this person on the left here has a low amount of cancer, and so and then this person on the right has a higher amount of cancer, and you’ll see, in terms of the colors, they have more genetic variability. So if you’re doing blood samples, when you have lower amount of cancer, might be harder to detect the cancer at all. But as you get more amount of cancer, it’s easier to pick it up, and then you also get a greater variety of genetic mutations in your sample. So one way to overcome this is to take multiple blood tubes, and you’ll see that commonly, if you have CtDNA detected, or if you’re getting that drawn, that you’ll have two blood tubes, as opposed to our standard one blood tube because they’re trying to look for these rare mutations.

Dr. Stacey Cohen 6:45
So to kind of summarize the section of what is it? CtDNA is an emerging technology. It can be very useful as the so called liquid biopsy, as long as you recognize the limitations of this type of test and the particular platform or company that you’re using. The detection is based on the amount of tumor in the body, and tumor informed panels may be more sensitive, especially for molecular residual disease. Though, again, as the tumor naive platforms get better, they’re probably going to be more head to head. Next, we’ll think about this as a marker of molecular residual disease. So this is the idea of MRD, which either could stand for molecular residual disease or minimal residual disease. It’s been published in both ways. And this was done quite a bit in heme malignancy, so blood based cancers, with the idea that you’re looking in the blood to sort of see if everything is gone. And so the idea in a solid tumor is that we typically use scans to try to determine if the cancer is there or not, but what we’re actually looking for is with MRD, is trying to find an even smaller volume of disease sooner than you would be able to pick it up on scans or other clinical measures. And the idea here is that CtDNA might be that biomarker that helps us evaluate for MRD and might be able to pick up recurrences faster than CEA or other blood tests. The problem is, is that most of our analyzes so far have been retrospective or observational, which is a bit limited.

Dr. Stacey Cohen 8:15
So let’s think about how do you figure out if a biomarker is a good biomarker or not? The two terms that we’ll use to describe types of biomarkers, one is prognostic and one is predictive. So prognostic means, does it help us better understand overall survival. And predictive means, does it help us better understand treatment selection so that’s specific acute action. When we look at different patient samples in a retrospective series, so this might be, let’s say, a clinical trial, where they happen to check some blood samples along the way. It had nothing to do with the trial, but they looked back to see if they could find a biomarker for how people did. It could suggest a potential role, but there could be other causes going on, and it could be the result of chance, because it wasn’t really part of the trial. specifically. When we collect things in a prospective or forward moving way or in an observational way, this is better. But you need to have a certain number of events or occurrences of something that you’re interested in to be able to statistically say that something is different between groups. So again, in an observational study, that’s better, but we still don’t have a clear answer. So the best way to figure out that a biomarker is the right thing to do to make decisions on, is that it’s prospectively collected, and you’re actively making decisions in the study based on that biomarker. So for example here, let’s say we had a CEA, and we said, if someone had a high CEA, we were going to start a new chemotherapy regimen. If that was built into the study, and that was true, that that seemed to help people, that would be the best option for that test to become a new standard of care. So with circulating tumor DNA, we have some retrospective studies, we have some observational studies. But we’re still working on these prospective type studies.

Dr. Stacey Cohen 10:05
So why do biomarkers matter? Because again, we’re trying to help at time point zero decide how we should treat patients to try to get the best outcomes. So this is an example of a Kaplan Meier curve, and so at the top, at time point zero here we have 100% of people alive, and in this case, recurrence free. And if we look at the axes down here, we’re looking at time. So if we just take the black line in the middle in this supposed patient population, this is how all patients did. But we know some people are going to do better than that, and some people are going to do worse. So the question is, if we had CtDNA was a good biomarker. If CtDNA was negative, could we actually give them and we knew that those patients did better, could we give them less intense therapy? Or alternatively, if patients are CtDNA positive and are more likely to recur, could we actually increase or escalate their therapy to try to make it more intense and see if we can impact those outcomes.

Dr. Stacey Cohen 11:05
So now I’ll start to talk about some of the data where we’ve actually looked at CtDNA and see how it would impact survival. So first we’re going to start with prognostic markers, so looking at overall survival. This was done within the IDEA France study, which was one of the three versus six month trials. This one done in Europe, looking at adjuvant FOLFOX or CAPOX and stage III colon cancer. So they had over 2000 patients in the study, but about 1000 had Ct DNA that was sampled prior to the initiation of chemotherapy. There was a slightly higher risk population compared to the total sample. They looked at two methylated markers by PCR. So remember, PCR is kind of, in some ways, our most limited way of looking at CtDNA. And this was a tumor naive approach. So everybody got the same two markers. They showed that 14% of patients were CtDNA positive after surgery, and that the disease free survival was worse for patients who were CtDNA positive, as opposed to patients who were CtDNA negative. When they looked at three versus six months, this was significant within the three months population. And when they looked at low risk versus high risk, this was significant within the high risk population. So when compared to that example curve, where the curves could either go high or low with CtDNA, they did separate out the two groups, but maybe not as profoundly so as the cartoon showed. So this platform here is low cost. It’s very reproducible. It’d be easy for someone else to run this test. But the how useful this is, checking these two methylated markers is less clear because we didn’t see as profound differences.

Dr. Stacey Cohen 12:51
Next, I’m going to show some data from the Australian group, which has published a lot about CtDNA. And they were looking also at how does clearing CtDNA and checking it over time, how does that affect prognosis? So this is 96 consecutive stage three colon cancer patients who had surgery and then they were planned to have six months of adjuvant chemotherapy. CtDNA was collected after surgery and then again after chemotherapy. And in this particular one, they looked at 15 commonly mutated genes and picked one from the tumor to track in the blood. So it was a tumor informed approach. The majority of patients, but not all of them got an oxaliplatin based therapy. In this test, 21% were positive after surgery, and that decreased to 17% after chemotherapy. But where I think this is most interesting is when you look at patients over time. 25% recurred overall, and the majority of these were in CtDNA positive patients. And again, the majority of CtDNA positive patients did have a recurrence, which typically happened about two months difference from what you would see with imaging. So looking here, we see a more polarized difference between patients who were CtDNA positive versus CtDNA negative, but it was even more polarizing when you looked after chemotherapy in terms of survival being much worse in patients who were still CtDNA positive versus patients who were negative. So when we look at this over time, this is where I think it becomes interesting. So if someone was positive after surgery and they stayed positive despite chemotherapy, the survival was quite poor. But if they were positive after surgery and then cleared their CtDNA and became negative, their survival was much better. Similarly, patients who were negative after surgery did the best, especially if they stayed negative, but if they were negative after surgery and then their ctDNA turned positive, their survival is much worse. So this shows that this is a biomarker which is helpful, not just at a single time point, but also over time, and helping us understand prognosis for patients.

Dr. Stacey Cohen 15:01
So how does this compare to CEA? Because we have another blood marker. Do we really need something better? In this particular situation, the CtDNA outperforms the CEA. And so we can see here, patients who had the CEA elevated did worse in purple versus the patients who are negative and in green for normal. But I think the interesting part is when you compare both CEA and CtDNA. So there were six patients that had high CEA and also who were CtDNA positive, and five of them recurred, so almost all of them. But when you looked at the discordant ones, there were six patients who were high CEA, but negative CtDNA, and only oneof them recurred, suggesting that the CEA was not the best predictor, and CtDNA may be better. If we compare it to when we would have seen it on scans, this is two different studies, this is sort of a funny curve, but you read it from the bottom, so time to detection of relapse. Here’s zero time point down here at the bottom. And in the blue dots we have CtDNA, and in the yellow dots we have radiographic. And if it’s pointed up towards the left, it means that it took a longer time to see it on scans. And a median of 8.7 months in both trials actually. In a separate panel where they did a tumor naive panel, the lead time was 11.5 months, suggesting that those tumor informed approach may be better.

Dr. Stacey Cohen 16:37
So of course, though, we don’t just live in a bubble, and we want to say, is there something that we could actively do to try to help patients clear their CtDNA? So then we look at, does giving chemotherapy actually help us do that? So in this study, we have 130 colorectal cancer patients, and 89% were positive before surgery, and 11% were positive after surgery. And when we look at that after surgery time point, the majority of patients who were CtDNA negative did much better than patients who were positive. If we look at how chemotherapy influenced that, approximately 70% remained positive despite chemotherapy, so only 30% were able to clear their CtDNA. But if they did, and they got grouped in this upper group, they tended to do better. So this particular tumor informed panel seems to help us make some decisions. But again, downsides of this type of panel is that it’s expensive, it’s proprietary, so you send it off to the company. We don’t really know which specific gene they’re testing. It requires tissue, but it could be highly informative and give us much more interesting information and helpful information than we might have had with some of the more rudimentary approaches from before. But the more that these panels improve, the better options we have.

Dr. Stacey Cohen 18:07
So this is very recently published data. 84 patients of colorectal cancer undergoing curative intent treatment, all stages and you can see there’s a variety of stages here. 45% in neoadjuvant, 54% had adjuvant therapy. So there’s a variety of treatment patterns. 24% were positive for their CtDNA, and when we look at recurrences, so 36% recurred. It was much more common in the CtDNA positive folks than CtDNA negative. And the sensitivity while low with that original panel, when they added epigenetics or methylated markers, then we could improve the sensitivity. In comparing that with CEA, as you can see here, it really wasn’t as helpful with CEA alone. So again, in this sort of newer version of a tumor naive panel, again, expensive and proprietary, but has the potential to be highly informative and have rare false negatives. So some folks have said, we should be doing this type of platform for most patients, because you don’t have to wait for the tumor sample and have enough tumor to biopsy, and it takes a little longer to get that first result if you’re using a tumor informed panel. But I will say that in looking at the trial itself, they still had quite a few patients that had some either sample quality issue or limited sample issue where they couldn’t run the test. So in neither the tumor informed or tumor naive is there a sure thing to do. So we leave it right now is that we have a couple of different options that we could do.

Dr. Stacey Cohen 19:47
So could we actually change our treatment plan based on the CtDNA? So again, this is going back to the Australian group. So this is stage II colon cancer, and patients had surgery. And they got adjuvant chemotherapy, as was recommended by their physician. But recall that many people in stage II, and probably most people in stage II, did not get chemotherapy. As expected, it wasn’t as common to have CtDNA positivity, but 9% had it after surgery, and it did reduce in patients who had chemotherapy. So if we look at their CtDNA status again, patients who were negative did much better than patients who were positive. And if we look at clinical risk factors, so like T stage and other surgical features, patients who were considered to be low clinical risk did better than they were high risk, but maybe not as profoundly separated. Where it’s interesting is if you take this low risk group and you say, is there any way to figure out, just becauseyou’re T3N0 does not mean you’re never going to respond, its just less likely. So if we look at that low risk group and break it down, patients who are low risk and CtDNA negative did really well, but patients who are low risk but actually had CtDNA detected, they did much worse. So this suggests that maybe we could actually figure out which of these stage II patients we should give chemotherapy to, if we were able to separate out, not just by CtDNA, but also by looking at our clinical risk factors and combining the two together.

Dr. Stacey Cohen 21:21
I want to point out that a lot of what I’ve showed has been colon cancer, but we do have very similar data in rectal cancer, just not as much of it. So this is an example of one of these studies where we have, now radiation included, and we’ll see that from diagnosis to the end of radiation to after surgery, we see a decrease in the detection of CtDNA. And if we get all the way to that point and someone’s after surgery, again, someone who’s CtDNA undetectable or negative at that point has much better outcomes than someone who has positive CtDNA. This particular study did not include TNT or total neoadjuvant therapy, where we’d be giving chemotherapy in the neoadjuvant setting or before surgery, but it remains interesting and good to know that these patterns exist in rectal cancer as well.

Dr. Stacey Cohen 22:09
Now when we’re able to be aggressive in stage IV, of course, we can sometimes take these patients through surgery as well. And then we would want to know if what we’re doing in earlier stage disease also applies to stage IV. So this is some data that I presented last year, which is a so called “real world” approach, where this was clinically ordered CtDNA. So again, observational data, without being within a clinical trial. But what we could see again is that for patients who had pre surgical testing, that CtDNA was commonly detected. And it was found still after surgery in about half of the patients. If they got chemotherapy, it was suppressed, but again in the surveillance time, so after the chemotherapy was done, we still saw CtDNA positivity in about 50% of patients, which goes along with the expected relapse rate. I haven’t mentioned it too much, but we can also quantitate the CtDNA or give kind of a number to that positivity. And as we would expect after surgery, the numbers would go down, and then when we’re starting to pick up on that CtDNA in the surveillance setting, it’s going to be a lower number, because we’re looking out for recurrences in a faster timeframe.

Dr. Stacey Cohen 23:24
This is a bigger study with that same assay. So again, looking at metastatic patients in Europe undergoing curative intent treatment. And they had their CtDNA checked at some defined time points, and a portion of them had adjuvant therapy. Again, about half of patients had CtDNA detected after surgery, and what they were able to show again is that recurrence was higher in the CtDNA positive folks, 98% versus 37 of negative.So again, just because your CTD negative does not mean that the recurrence is zero, but it does suggest that it’s a much lower risk. If we look at a single time point after surgery, we see that there is good separation of the curves, and CtDNA negative patients do better than patients who are positive. And if we follow with multiple subsequent time points, then we can really tell that better over time. And I put some of the information here. So before we were talking about a lead time from imaging of 8, 9, 11, months, because these patients were being imaged a lot more often being stage IV, the lead time was 3 months, but still preceded what we would expect for radiographic occurrence. So to summarize this section on MRD, ctDNA positivity at the end of treatment is associated with worse outcomes. Detection of CtDNA during surveillance is highly associated with recurrence and may precede what we would see on imaging by several months, but we still need prospective data to use CtDNA to adapt therapy recommendations.

Manju George 25:04
Dr Cohen, can I ask a question? This is excellent. Lots of great information. So my first question was about the rectal cancer trial, where you said that they were doing radiation and then using CtDNA to show that after every step, the CtDNA levels were coming down, right? My question was did they mention in the study,what is a good time after radiation therapy to do CtDNA testing?

Dr. Stacey Cohen 25:32
Yeah, I think it remains a good question. I don’t think we fully know the answer. So after surgery, we typically say two to four weeks would sort of be the ideal time to check that. And there’s been other studies trying to find that key time point. It’s a little harder after radiation, because you don’t just take the tumor out of the body and then the blood can clear. Here, there’s a process of cell death that can last six eight weeks, so I don’t think we truly know, except to say that very few people are willing to wait that full six to eight weeks to find out just on a blood test, what’s going on. So I would think checking it around the four to six week mark might make the most sense. But in reality, we could be checking it sooner. I think we just don’t have the data for different time frames to understand when the best time is.

Manju George 26:20
Okay, thank you. And then my other question would be about what you just showed the stage IV treated to Ned, so the right side graph said that 37% recurred even when they were CtDNA negative. So was that like one time testing?

Dr. Stacey Cohen 26:40
They did a couple of time points, but again, it was still kind of an observational study. So we don’t know if that would really exactly mirror what’s currently done in clinical practice, but they did have a few time points that were checked there.

Manju George 26:58
Okay, so in relation to that, is it true that, like, for example, with early stage they say that the the highest chance of references in the first 18 to 24 months after the end of treatment. Does that hold good for this group, stage IV treated to NED?.

Dr. Stacey Cohen 27:13
Yeah. So I mean, many of the CtDNA studies are only checking it for two years after the time of surgery, regardless of any chemotherapy or anything else that’s given with the idea that if by the end of that two year time frame, you’re still negative, you also have this lead time, hoping that that would be farther than what you would see on imaging. And so two years should be sufficient, because the majority of recurrences do happen in that early time frame, and probably happen even faster for stage IV.

Manju George 27:43
Okay, thank you very much. Yeah. Does anyone else have questions? Or can we go forward? Yeah, I think so.

Dr. Stacey Cohen 27:53
Great. Thank you. I just wanted to throw up here a couple of studies, because of again, it’s really important that as we learn about CtDNA, that we’re supporting the efforts to learn more. So we really can make this part of practice, if that’s the right thing. And so I just included a couple examples of studies. There’s not a lot out there, but there’s a couple of examples, so we can both think about de escalating or escalating therapy. But the ones I show here are largely looking at, can we give different or more intense chemotherapy for patients who are CtDNA positive? And I expect that this is something that is going to be very well studied in the future, moving forward.

Dr. Stacey Cohen 28:35
So lastly, I’ll talk about the use of CtDNA in active cancer patients. So we already talked a little bit about a tissue biopsy versus a liquid biopsy. So the liquid biopsy is certainly less invasive, has less cell content, and there are pluses and minuses with it, but it is relatively easy to obtain. Does it actually matter, though? So this is an example of of why we could still use it. So this is an example of metastatic colorectal cancer patients who are receiving first or second line therapy. 77% had detectable ctDNA at baseline. Now this is lower than some of the other studies, and again, it also refers to what type of panel you use. And I think a bigger take home point is that it really matters which panel in which study, because we can’t necessarily say that we learn from one study to the next. And then, you get one company and then the other company, because each one sort of has to be replicated on its own. But within sort of the general concepts they seem to be the same. So what they showed is that higher levels of CtDNA were associated with worse survival. Which sort of goes along with what we’ll see with CEA, so maybe not that different. We can use it to detect commonly altered mutations. So KRAS, BRAF, these are key molecular markers that we need to know about every metastatic colorectal cancer patient. In this particular study, all had detectable CtDNA, and it had very good accuracy when we compared tissue to the blood based assay, so showing that these were reliable ways of checking it. But a more interesting thing that we’re starting to think about is, could we use this more longitudinally? Could we do this over time, to try to think about re exposure and whether or not that’s working? So as an example of this, this is the CHRONOS study that was just presented at ASCO this year. Patients who had prior EGFR therapy, not right before, but at least two lines before. So think of it, maybe someone who had it in first line, then they added different therapy in second line, and now you’re trying to decide what to do for third line. They used digital droplet PCR, so again, our more basic method, on cell free DNA to look for resistance mutations. And if patients seem to still be RAS RAF wild types, suggesting that they might respond to EGFR therapy, they were offered to go on this re challenge study where they received EGFR inhibitor again. And the response rate was 30%, they said they expected a response rate if we hadn’t done this test of 20%, so maybe it was a little bit better. And disease control rate of 63% with a progression free survival of four months. But what it concerns me a little bit is sort of what we see here over time. So this spider plot, what it’s showing is how patients did over the course of the study. So for lines that bend up, those are patients that are progressing. And in red, we have patients who cross the threshold where they’re going to call it progressive disease. In gray, you have the patients who have stable disease, and then in green, patients who actually hada response. And so you can see here that there is a group of patients that seem to respond, or actually have at least 30% tumor shrinkage. But those responses don’t seem to last very long, so almost as soon as they go down the lines go back up, suggesting that those responses were not durable or not maintained. So it means that, yes, we can potentially re expose patients to older therapies, but it still always begs the question of, did it work and is it worth it? And I think that’s something we’re still trying to figure out.

Dr. Stacey Cohen 32:19
One nice thing about CtDNA is we can use it to detect rare fusions. And this is important because sometimes there are targetable mutations and therapies like NTRK. And might work better than than tissue in certain examples. So for NTRK, specifically, if we look at it here, they were able to detect it at a slightly higher frequency as compared to tissue. But again, if we look at the scale here, these are incredibly rare. And notably, there was a lead time or difference of 24 months, (two years), between when the tissue was collected and when the CtDNA was collected. So if you think that someone’s mutation profile may be changing over time, going back to the old tissue and trying to find a mutation may be less helpful. But of these, the only one really approved for colorectal cancer is NTRK and they’re very similar. So I’m not sure that we necessarily need to all go out and get retested, but I think it’s just important to know that this could work quite well.

Dr. Stacey Cohen 33:21
Microsatellite instability is a lot harder because it’s not one specific genetic marker, but you’re looking at this pattern of change in the tumor. And so when we look at CtDNA versus standard methods, it seems to work pretty consistently. When we compare it to PCR and NGS, there are some discordant cases. So these are the ones in blue here where the CtDNA was positive, but the tissue was negative. And it probably has to do with the fact that IHC can sometimes be a little bit buggy, but in general, it was very well concordant, which are these dark gray and red bars that you’re seeing here. Again, we can use this over time. So this is an example of monitoring CtDNA while someone’s on therapy. So these patients were getting PD-1 therapy. And so this was from not just colorectal cancer, but other cancers as well, as long as they were MSI and getting this PD-1 therapy. And when we look at progression free survival, we see some differences.Looking at the tumor mutation burden in the plasma, patients who started out high did better than patients who were lower. And interestingly, you would expect all these patients were MSI by tissue, and so that’s how they got enrolled on the study. But what this was showing is that if we actually looked at the tumor mutation burden by the cell free DNA, that was probably a better marker for tissue response than looking at old tumor tissue for MSI analysis. So maybe it’s because it’s real time. Maybe it’s because there’s some variety what they call that mosaicism, where there’s different patterns of mutations in different parts of the tumor. And that kind of gets equilibrated out when you take a blood based test, but it’s interesting and thought provoking. Again, we right now don’t do this in trials. We don’t know if this is really the right or wrong thing to do, but it is interesting that maybe this could be something we would start doing.

Dr. Stacey Cohen 35:18
So finally, to summarize, so non invasive, liquid biopsy can be helpful to assess the cancer profile, or if a tumor biopsy can’t be obtained, then we could definitely use that to do molecular markers. Some platforms, but not all are validated for MSI or tumor mutation burden. And so these are things that, of course, if we were checking that we’d want to know in advance if our panel could do that. And importantly, if we take these tests over time or serial biopsies, we can sometimes watch how someone does, either from resistance or success, but the problem is we don’t know how to apply this to the clinic. And there remains issues regarding cost and reimbursement. So where are we at today? With CtDNA, these tests can be ordered by your physician. We can debate tumor informed versus tumor naive type panels, and there are good options and bad options for both, and there’s pros and cons of each strategy depending on the platform. So it would be deciding with your physician which they’re comfortable with and what you’re kind of looking for. The problem is it’s not yet clear how treatment decisions should be changed based on CtDNA results for minimal residual disease. So at this point, I consider it as another leg to stand on. We have our clinical factors, we have our imaging, CEA, and now we have a fourth leg to that stool, which is CtDNA, but by no means is it the be all and end all. It’s also not currently part of the NCCN or national guidelines, and I highlighted some of what they put here, which is that we right now don’t have that predictive data. We don’t know how we should use the CtDNA results to absolutely change therapy, and so because of that, we don’t yet recommend this for every patient, but the NCCN encourages enrollment in clinical trials, because we think that at some point this may be really useful information. And with that, I’ll say thank you for listening.

Manju George 37:09
Yeah, thank you so much. This was great. I have a couple of questions. So your slide on NTRK fusions. You thought that both the tissue and CtDNA were kind of similar, right? So do you suggest, then that for people who are interested, it’s best instead of having a new biopsy or tissue biopsy, it might be just better for them to do a blood test, right?

Dr. Stacey Cohen 37:34
Yeah, I think it would be fine to do a blood based biopsy rather than going to the old tissue if, it was that or get a new Bnew tissue biopsy, it might be very reasonable just to do the liquid biopsy instead.

Manju George 37:49
Okay, my other question was, like, I was kind of interested in what you said about that figure for MSI high, where you had that blue box, where there was some discordance between tissue IHC and CtDNA. So from your personal experience, do you think that the CtDNA test is better than the IHC? Does IHC have more false results? Is that what you meant?

Dr. Stacey Cohen 38:14
Right. So IHC, even though it’s one of the most common ways that we look for MSI in the tumor, it does have some errors, because we can’t always tell, it can show artificial loss of expression at the wrong times. And so my institution, for example, pretty much only runs next generation sequencing because we think that it gives the highest quality result. And so this would be next generation sequencing on the tissue. And I would say that versus the liquid biopsy are probably pretty similar. But it’s a more complex test to do NGS, more expensive, and so depending on institution and resources, IHC is still done routinely, but there sometimes is some interpretation that’s needed for it.

Manju George 39:07
Okay, my last question was about CtDNA for MRD. What you mentioned was the NCCN guidelines and all about escalation or de escalation of therapy. What is your opinion about using it for surveillance post treatment?

Dr. Stacey Cohen 39:24
Right. So I think that different providers have their different biases about whether or not they think that this is potentially a useful platform. Clearly, I have some bias thinking that it could. But I also remain cautious, because I’ve had patients say, start chemotherapy in the adjuvant setting, let’s say for low risk stage III, and say, okay, my CtDNA came back negative, can I stop? And I’m not yet comfortable making that determination. So I say it just gives us more confidence in maybe what we’re planning to do anyway. So if we were thinking, okay, you have a high risk stage II, and you really wanted to stop at three months, because we think that the differences in survival are only a couple of percent there. Maybe that CtDNA would give us some confidence, if it was negative, to stop early, and if it’s positive, maybe that would give us some motivation to say, let’s really try and clear it with the six months and see if that helps. So I think that it can kind of add to what you’re going to do anyway, which is why I think of it as another leg to stand on, as opposed to the only thing that’s determining how we treat patients.

Manju George 40:36
So but what about like, two years after adjuvan therapy? Would you recommend that among all the other tools, like, forexample, a patient who’s CEA is not useful. If it was like one or less than one, what in your mind would be the value of using a CtDNA test for surveillance?

Dr. Stacey Cohen 40:55
So I like to use it for my patients. Again, not everybody does, but I have been checking it on a lot of my patients. However, if someone was already a year two years out, I probably wouldn’t start checking it now, because I think in terms of timelines, the majority of the recurrences do happen earlier on, so I don’t know how much benefit we would getin starting to do surveillance testing at this point. But if someone was newly in the last few months, had had their surgery, then I think would be reasonable to ask your provider about if this was a test that would be useful.

Manju George 41:32
Okay, okay. I think we have one question in chat. Yeah. So he said the heather, do you want to read your question?

Heather 41:44
Yeah, sure. Thanks. So unfortunately, we just haven’t been able to get any traction with getting my husband’s care provider to even really have a conversation with us around potentially ordering a CT DNA test and the unfortunate thing is, his scans, nor his CEA level have given us any insight, and so I feel like this could be potentially beneficial. And when I had the conversation, I said, I don’t necessarily want to do this for you to drive treatment considerations off of but I think informationally, it may potentially give us a heads up. And I mean, they just shot us down quickly and wouldn’t even… So I mean, is there something I could say differently, or a way I could phrase the question or go about it to maybe get them to entertain the idea?

Dr. Stacey Cohen 42:34
Yeah, I think it’s really hard. Because, again, I think to some degree, like anything new, there’s kind of believers and non believers, and it might just be that the physician that you’re currently working with just doesn’t feel that it’s going to be reliable, or doesn’t think that we will know how to interpret the information well enough. And I mean, there’s a good lesson in medicine, which is, don’t order a test that you don’t know how to act on, right? So, but at the same time, I think that you named some really important features, which is when if we’re standing on those four legs, and we can’t stand on the CEA one, and the imaging is a little iffy, then it’s a lot harder to do surveillance reliably. And so, I feel like I might, not knowing the full story, I might be comfortable ordering in that situation, but a colleague might not. And so maybe one thing to do is just to consider a second opinion of someone at an academic institution nearby, or if you’re at an academic institution, a different one, and see if just a different provider might be more open to it, because I’m not sure you’re going to be able to change someone’s mind, because, again, the should part is not there. But I can tell you, anecdotally, I’ve had number of patients where the regular scans were unrevealing, the CtDNA went positive or high. And then we did a different type of scan, like an MRI or a PET scan, because of their unique situation, and we did detect a recurrence. So I have had situations where I do feel that the CtDNA helped me, and often it was in patients where the CEA was unrevealing, because not everyone makes CEA, so we don’t always have that as an option.

Manju George 44:20
Thank you. Thank you. Yeah. And what we have done in COLONTOWN is that since people started using Signatera, we have a Google Doc, or a Google sheet where any person who has ever had CtDNA testing would write the name of theirdoctor and this cancer center where they got the test. So our idea was that, when you have a new patient coming in and who’s interested in Signatera, we would say, hey, take a look at the list. If you have a doctor in your institution, somebody else who has ordered the test, the simple thing would be to tell your oncologist. Could he or she talk to that person ? Maybe that would change their mind. Or if you see the list and then see that they’re not the only person in that whole state using it, but there are a couple of people who have used it, maybe that would make them want to learn more. And kind of be helpful for the patient. The other question that I had is, like, recently, we have seen some plasma TMB tests that people are getting, and then the values are, like, 30 or 40 or something like that, when compared to, like tissue TMB, where the FDA approval for pembro, for example, is at beyond 10, right? And then there is some confusion, because, I think that there was maybe one paper or something which said that a plasma TMB of 18, maybe or 28 I don’t, can’t remember it right now, that was somehow similar to a tissue TMB of 10. I mean, I see these kinds of comments where people are confused about how to interpret it, and especially when they see the numbers, they’re like, Oh, should we be doing immunotherapy now that we have the number?

Dr. Stacey Cohen 46:08
Right. And I mean, I would say, even with tissue, we think that people are unlikely to respond at 10. With colorectal probably need 12 or higher so, and maybe even 15. So even though the FDA might make these blanket statements within an individual cancer, it might not be exactly correct, so I have to look at the specific test. But again, I think this goes back to that issue that even though we think of CtDNA as one concept and tissue is another concept, each individual person running the test may report out or analyze the data differently, such that we can’t always apply it across different ways of looking. And so we’d have to look and see like, kind of what their scale is in that particular plasma assay for how they call higher or lower, and then try to kind of equilibrate that down to what we would do for standard tissue based testing. Yeah, but yes, it’s incredibly confusing. Yeah, which is why I think a lot of doctors are not ready to order it, and that’s okay, because, again, if you don’t know how to interpret the results, as many providers do not, then it makes sense to not order the test that you might not feel comfortable with. But I think that this is a technology that is not going away anytime soon.

Manju George 47:24
Okay, thank you so much. This was great, and I’m hoping that I can edit it and post it in CTU so you can take a look and you’re welcome to share it with your patients, or you know however you want to use it.

Dr. Stacey Cohen 47:35
Okay, sounds great. Okay. Thank you. Bye, everybody. Thank you.

Manju George 47:38
Thank you! Bye. Thank you all for attending.