Molecular Residual Disease (MRD) Testing in Colorectal Cancer (CRC): Adjuvant Setting (Round 2)
Video Transcript
Dr Namrata (Neena) Vijayvergia: All right, so here we go. Now, the next part we're going to talk about is the adjuvant setting, right? So, we already discussed the neoadjuvant, which is before surgery. Now, we are moving to—patients have had surgery done. They've resected colon cancer. And we are deciding if we should give them adjuvant therapy or not. That's where the most of the data in this disease is, tons of data that's out there that we can talk about. So, here is the setting. Now we are looking for a tumor that's not visible. So, it's gone below the detection line. And we're trying to identify if we can help these patients because we know some of them will recur and cause problems. So, the way I want to think about it in the adjuvant setting, for a test for me to—for a test to help me, I want to look at the assay itself. It has to be a good one, right? Reproducible, something good that I can use every time and everybody can use it. I want to make sure that I have enough data in the clinical scenarios that I'm going to use it. And I also want to make sure I look at the patient perspective, right? If I tell somebody you're ctDNA positive, what happens to them? Do I wreck their life? Do I, you know, make them empowered? Like, what happens? So, we want to look at all of these areas. So, first of all, off the bat, ctDNA is a very good test to prognosticate a patient with colon cancer, right? If you had surgery, multiple studies, and you can see here the percentage of—if you are ctDNA positive after surgery, as we go down by the year, the sensitivity, like, so the proportion of recurrences in a positive patient increases, which is just suggesting our tests are getting better, right? We are improving the tests. And as a result, like, by the time we come to the GALAXY study, which is the Kotaka study, 61% was the rate of recurrences in patients who had positive ctDNA. So, very high risk. So, it's a very good prognostic marker. But does it end there? Can we cross that bridge from prognostic to predictive? Now, as we come to the test, right? Any test you do, right, if a doctor, if I as an oncologist—I'm a community oncologist. I'm seeing different types of cancer patients every day in and day out. And a new test comes in, and I say, you're going to use the ctDNA. They're going to ask me, when do I order it? What is the best time to do it? So, that's the problem. Like assay, we have to standardize it. DYNAMIC study, which is— we're going to talk about—which was a phase two study for supporting its role. It had two samples done at week four and then at week seven. The COBRA study, which we don't talk about much, but it was another study, which had a single test between two and six weeks. The GALAXY study, the Japan study, which is a prospective study in stage III patients, did one sample. The four-week postsurgery is used to show the results and benefits of adjuvant therapy. And then, there's the BESPOKE trial, which is the American version of the GALAXY trial. You can say it utilized an MRD window. You could do up to three tests in the 12-week period. Now, tell me, if I'm a community oncologist, which of these should I follow? We don't know for sure yet, right? You could do one or the other. So, we need to optimize that. We need to come up to a conclusion as to when are we going to test our patients? The turnaround time is another pet peeve, right, of this test. So, the test, we all think that tumor-informed assays are the ones we want to use, right? Because the amount of data with the tumor-naive assays is not as much at this point in time. The turnaround time for these tumor-informed assays is four to five weeks. We have one prospective study, the DYNAMIC trial, which used this tumor-informed assay at two different time points. The time to start of chemotherapy in that study was 84 days. So, the day you see, like, it took 84 days from your surgery to actually get chemo. Now, that's like, I feel like sometimes you're sitting on a ticking time bomb, right? You don't know—how can you not treat these patients? There was a meta-analysis a long time ago, which showed us that delaying chemotherapy after eight weeks is detrimental to patients. So, that is another area we want to work better on, an area to improve on. The DYNAMIC trial, right? This was the study that, you know, we had heard the DYNAMIC-Rectal before. This is the DYNAMIC trial. This is what we used to guide our decision for stage II patients in the adjuvant setting. Here, what they did, very simple. Stage II colon cancer patients have surgery. Positives are CTD—so, there was half of the patients got treated as we normally would. Other half, the medical oncologists were told their ctDNA results. And then, they say, decide what you want to do. So, it's a ctDNA-guided management versus not, or standard management. And what they found here was more patients received adjuvant chemotherapy if we follow the standard regimen. So, less people got chemo if we actually did a ctDNA-guided approach. Good thing, right? Really good thing. We also found, that because we found a positive ctDNA, we just ended up giving more doublet chemotherapy than single agent. So, typically, in stage II patients, we do 5FU. But here, we did 5FU oxalipatin or something because we were nervous what ctDNA positive meant. We do not know if actually it is the right thing to do. We do not know that yet. This is how it was done. More people got it, so more neuropathy in those patients. And the recurrency survival was similar when the patients got, you know, equal. We have since then updated—the recurrency survival at five years is down to, like, 88% in the patients who are treated group versus not. And the five-year overall survival was about 93% in both groups. So, we know that this approach sort of works in this. But let me point to you. The T4 tumor patients, so patients who are high-risk stage II and T4 in the ctDNA-negative group actually had an 18% to 20% rate of recurrence in the ctDNA-negative group, too. Because why? Because we know most of these recurrent peritoneal recurrence, that's very low shedding. As a result, you don't see—catch many of these patients. So, if you decide that, oh, I'm going to use this for all my low-risk patients, but I'm going to also give all my T4 high-risk patients chemotherapy, that the improvement in the number of patients getting chemotherapy that I showed you, 15% versus 28%, goes away. You're getting more chemo to people. So, again, it's a really good number, but when you look deep, not sure that actually holds true. This is the GALAXY trial, right? The Japan—this is basically a prospective study where patients with stage II or III colon cancer were followed with lots—you have, like, now we have updated data of 2,000, over 2,200 patients in this group. And what they looked at, you know, you're going to get a bloodwork done at four weeks post. And then, they're followed for future. And then, if you become positive, we have a trial for that. If you remain negative, there's a trial for that, really cool design. In this GALAXY trial, very prognostic. We already showed you, right? If you're positive, you have a very high rate of recurrence, if you're negative. The second one, for some reason, my header cut off, but it's basically the outcomes of patients who got chemotherapy in the stage II and III. And you see that the patients who get chemotherapy and are ctDNA positive are—so, the red line is who didn't get chemo and the black line is who got chemo. So, if you're positive ctDNA and you get chemo, you have way better survival than if you don't. So, that's really good, right? But this is a, you know, not a randomized study. It's just prospective evaluation. We did see, though, that the rate of ctDNA clearance was very, very high in this group. So, with adjuvant chemotherapy, the rate of ctDNA clearance went from 60% to 10%. Like, so, ctDNA positivity after chemo went from 60% to 10%. We know that our chemo really is able to cure about 20%, 25% of the patients. I don't, and why is it getting so much better here? It's probably because of only something called transient clearance, right? That they get better, and then it comes right back. So, you have to improvise our test to make sure we identify those patients and understand what to make of it. The single test sensitivity of the test here was only 48%. Compared to where other studies did two tests or three tests, there your sensitivities was much higher. So, again, those raise concerns about the assay itself and the data that we have. This is the BESPOKE trial. BESPOKE is an American version of, like, a prospective study of patients getting their results and us deciding, based on the results, if it helped us. It's not telling us what to do. It's just giving you the results and letting you decide what you want to do for those patients. Again, you see the same thing in patients who are MRD positive. The chemotherapy made a difference. You can see the difference in the curves. If you were MRD negative, so the ctDNA was negative, there was no difference in the curve, right? So, prospective, but collection of data. But the results—but the treatments were not standardized. So, looked interesting. But they, you know, they had three tests, up to three tests in a 12-week window. So, the sensitivity of the test was much better. Can we actually do that in a 12-week window? I cannot even get one test done for patients sometimes. That's something to remember. This was an updated analysis of the BESPOKE study. It was just presented at ASCO GI. And really, did you see? They asked the physicians, did the results influence your treatment? 83% said no. So, is it ready for prime time? We have studied the BESPOKE study, and 83% patients saying that it didn't affect their decision. And it's not because they don't trust. It's a very good test, extremely good. But we just don't know how to use it very well so that assay can actually affect the treatment paradigm quite yet. At the end of the day, the patient's perspective is important, right? BESPOKE did a very good job at getting patient perspective on trying to find out, if I give you the results of ctDNA, does it make you feel better? Does it make you stressed out? What happens? Really, 92% to 90—like, over, really a high number of percentages you're seeing, they're saying it's good. We want to do it. We want our results. It helps us. I think partly because a lot of these tests were negative. So, if you have negative tests, it reinforces that everyone's going to be happy. Like, yes, I got my results. It's really good. But at the end of the day, it's like, I don't want to be an astrologer looking at the stars and telling you that this is your prognosis, but I have nothing to change it. So, I don't want to be that person. And I think, unfortunately, that's where we are right now. And I want to point out this very interesting theory that's outside of medicine. It's this diffusion of innovation theory that people who adopt new tests early on, are the ones who are already excited about that thing, right? Who got the first Tesla? People who believed in it. It doesn't represent the feeling or the emotions of the general majority. So, these early adopters is what we are seeing the results of in the patient perspective. We really need to get more data to see when it's used out in general population or the patients who end up being positive and we didn't do anything for to get more data to understand this better. So, ultimately, I do want to point out this study that there is hope, right? This is all we know thus far, but we are doing things to make it better and understand it better. This was the CALGB 80702 study in which—it was a very good trial. At one point, we were trying to see if adding celecoxib improved outcomes in colon cancer in the adjuvant setting. The study was negative overall. But then, they did the ctDNA analysis. And they found that patients who are ctDNA positive, there's a significant improvement, as we can see here. So, this is a, you know, it's a retrospective evaluation of a prospective study, just pretty interesting and something to consider. And then, we have a CIRCULATE-US going on to help us identify the answers.
Dr Jason Zell: Okay. Thank you so much.
Off-Camera Question: You talked about MSI high. It's uncommon to see MSI high in the clinic. You talked about rectum. That's less common in the clinic. Let's talk about the common patient we see. Last year during debates, the speaker for Cornell said—because you mentioned eight weeks—in order to make the decision by week eight, you have to have had a surgeon who draws the blood prior to surgery. And that's not—I guarantee you in the community that never happens. So, how do you resolve this at a stage II when you've made that decision by week eight because—
Dr Namrata (Neena) Vijayvergia: I agree with you completely.
Off-Camera Question: But that's the most important question.
Dr Namrata (Neena) Vijayvergia: That's the timing of the assay and the time it takes to get it.
Off-Camera Question: So, how do you resolve it?
Dr Namrata (Neena) Vijayvergia: I think we improve the sensitivities of the test.
Dr Jason Zell: I think that comes up in my section right here. So, I’m happy to cover that. Thank you, it's a great question. So, all right.
Dr Yelena Janjigian: We have representatives from Natera here who we can put on the spot.
Dr Jason Zell: Sure, you can put me on the spot for now and then go to, yeah, the rest.
Off-Camera Comment: I'm happy to catch you up on how we’re solving this.
Dr Jason Zell: So, I will—I may or may not win this debate, but I'm going to declare the winner of the best slide in the deck to Dr. Vijayvergia. I mean, that one with “I don't know” in the constellation. That was Pashtoon Kasi from Cornell at the time. I had the same feeling when I was sitting in the meeting. So, yeah, I might borrow that slide for the future. I don't know what to do with that. Patients are negative, they're going to be happy. Patients positive, they're not so happy, right? So yeah, it's more complicated. A more nuanced analysis would be necessary. Okay, here, this is what we're talking about. Patients have surgery, exactly your point. What do we do next? This is the adjuvant setting. All right, so let's come down from the constellation, right? Let's look at where things are today. And I will just say plainly, MRD assays are widely available. They're highly valid, analytically, clinically. And they have a high clinical utility. And there's been exponential increased use in the US. One thing I'm not hearing from my opponent or any opponent is, where is the phase three trial of inadequate lymph node sampling, given FOLFOX or not? Or even pT4 by itself, where's a randomized trial of that? I haven't seen it. And how about poorly differentiated tumors? Nobody talks about that. And we used those for years to decide who gets chemo, who doesn't in stage II, right? And we still do.
Off-Camera Comment: But, in 1970, 12 nodes or less was common. In the present time, I have not seen one patient who didn’t have that [inaudible].
Dr Jason Zell: Yeah, true. But that's a marker of the surgeon and also the pathologist at your hospital doing a better job of counting. So, but no one's going to do a study like that is what I'm saying. And so, what we had is prognostic information, and we made chemo decisions based on that. What we have here with the current assays are an extraordinary amount of robust data in the prognostic setting, and the predictive data is coming. So, just like we talked about in the neoadjuvant setting, this all boils down to the personalized approach. What we want to be able to do is either dose escalate for those patients who need it or maybe de-intensify therapy or no therapy after surgery for those who don't. And this was the best trial we had. We talked about it briefly, DYNAMIC study. I'm going to just describe this much more simply. So, yes, two-year recurrence-free survival is the same if you use the ctDNA-informed approach where they had—15% of these stage II patients had ctDNA positive. RFS is the same if you let the investigators choose how they want to treat or do this informed approach. But look, only half of the patients received chemo. And yes, more of this half received oxaliplatin. Maybe those are the patients who needed it, right? So, I don't know. RFS is the same as overall. I think these data hold up. I would also look at this. So, this is the GALAXY study in Japan. If you just restrict to the 2,200 patients that have stage I to IV resected cancer, there is a, like, whopping difference between ctDNA negative and positive. I mean, this is an extraordinary difference. More than what you would see for any of those pathologic features that I mentioned. If you look at event-free survival, 9% versus 58% events alone, or 89% DFS versus 28%. It's an extraordinary difference.
And now, we're getting into this area of, either in the MRD window for DFS and then OS on the bottom two versus during surveillance, if someone converts to ctDNA positive. These are extraordinary differences. A hazard ratio in surveillance, if someone becomes ctDNA positive, is 33-fold worse outcomes compared to somebody who's DNA negative. It's impossible to ignore the prognostic import of that. So, two-year DFS, 92% versus 6%, 8%. I mean, there's an extraordinary difference. So, from the US, this is the Signatera-based assay and the final annotated cohort with people who had blood draws in the MRD window, 154 patients. Dr. Dasari called this, the first time I'd heard it, Pelican beak curves. And I really like that term because look at the shape of these curves. It looks like a pelican opening its mouth. It's such a difference between the CT-negative group and the CT-positive group. So, I just think, yes, is this a game changer? Absolutely. These are the best data we've seen for prognostic setting. Now, what's going to answer this question, and this goes to your question right here. We have Arvind Desari and Christopher Lieu from the NCTN. They're conducting this trial of CIRCULATE-US where patients will first be identified, stage III colon cancer patients, ctDNA positive or negative. If they're positive, they'll be randomized to standard chemo or intensification. In this case, it's FOLFOXIRI. If they're ctDNA negative, they'll be randomized to standard chemo or surveillance. And that might be the most controversial arm. But it's also the most exciting because we know chemotherapy only helps cure a very small percentage of these patients, right? So, this is an exciting trial. What does this trial do? And this is in answer to your question. It allows a longer period of time for patients to get their ctDNA back. You're right. Eight weeks is the goal. That was the goal.
Off-Camera Comment: You didn’t mention stage II. Stage III we don’t worry about [inaudible] because we offer them chemotherapy. Even if their ctDNA was negative, we would highly consider supplementing. Stage II is the problem.
Dr Jason Zell: I think it's the same here. Like these people will never get chemo. So, these people are going to be treated just like—
Off-Camera Comment: —[inaudible] can’t make a decision by week eight?
Dr Jason Zell: Correct, you can't. That's exactly right. You can make it by week 10, which they allow in this trial. I will say it does—it's difficult in the community, but the import is on getting early surgical referrals. I think it's hard to get the blood pre-op, and I'm not sure the company will process it without the tumor sample anyway. In a very active colorectal surgery group, you can get biopsy tissue and blood pre-op, but I know that's rare. So, yeah, my goal is to get patients within one, two weeks post-op. And at that point, it is very much feasible.
This is a non-CME activity. The views and opinions expressed by the presenter(s) do not necessarily reflect the views and opinions of the Oncology Learning Network, HMP Global, or its employees and affiliates.
