Circulating tumor DNA (ctDNA)-guided adjuvant chemotherapy de-escalation in the treatment of Stage III colon cancer from the ctDNA-negative cohort of the DYNAMIC-III trial

Tie J et al. ctDNA-guided adjuvant chemotherapy de-escalation in stage III colon cancer: Primary analysis of the ctDNA-negative cohort from the randomized AGITG DYNAMIC-III trial (Intergroup study of AGITG and CCTG). ESMO 2025;Abstract LBA9.

Bekaii-Saab T. Discussant of LBA9. ESMO 2025.

Tie J et al. Circulating tumor DNA-guided adjuvant therapy in locally advanced colon cancer: The randomized phase 2/3 DYNAMIC-III trial. Nat Med 2025;[Online ahead of print]. Abstract

DR LOVE: So let’s talk about the DYNAMIC-3 trial presented at the recent ESMO meeting got a lot of attention. Dr Saab did a discussion of the paper. We see the publication now in Nature Medicine just came out. So can you talk a little bit about this study and what they saw?

DR KOPETZ: Right. So this is the Australian — or the ATITG study that was looking at the Stage III population, kind of a Phase II power. The randomization’s a very practical study to standard management, where ctDNA results weren’t utilized, or, in a Stage III patient, if you’re negative, to deescalate what you’re planning to do, and if positive think about how can you escalate treatment.

Now in Australia there is a bit of monotherapy 5-FU that was utilized here as part of standard management, but most patients were getting the adjuvant with FOLFOX. We did see less chemotherapy utilized, so that again, the ctDNA guided saw an overall lower rate of adjuvant therapy there. This importantly was a single timepoint, so postop single draw with you would argue maybe a bit of an older but still tumor-informed ctDNA assay and used that to make that treatment decision.

What was seen here were curves that were close, but there was a little bit of a separation there, and they had a noninferiority margin that they were not able to demonstrate noninferiority in this with it.

So what we saw was that if you’re ctDNA negative you had a very low risk and that treatment deescalation was feasible with high adherence, with a lot less oxaliplatin utilization, better outcomes there, but not able to confirm with the barriers to the current assay the standard of care was noninferior — or that ctDNA defined was not inferior to standard of care.

DR LOVE: Anything you want to say about Dr Saab’s commentary after this was presented?

DR KOPETZ: No. I think it highlighted the reality that we have to make decisions on the basis of our — of the data we have in hand and that while we want prospective data to help guide us we have to really consider clinical — the significance beyond the statistical margin, especially for these smaller studies like the DYNAMIC-3.

So the point about you’ve got to have a good test and the idea that it’s not perhaps a single timepoint test, but serial testing is the other kind of component, as mentioned, that is really critical.

DR LOVE: So you mentioned, I guess, this was another bespoke assay, but it was not Signatera. Any thoughts about is there a difference in quality between these? Could that be explained, I guess, the lack of — the fact there was a slight difference there?

DR KOPETZ: Absolutely. I think there’s — still we think tumor informed are the kind of best type of tests. We always have to execute the study with the best assay that’s available at the time the study launches. If we were kicking off a study now would it be that assay? No. But we do have better assays that are available now that allow us to better define that population maybe that we don’t want to deescalate.

DR LOVE: So a lot of this discussion kind of reminds me of what happened in breast cancer when the Oncotype came out in 2004, and that became a real critical factor in decision making in oncology and breast cancer. And one of the things that came along with that, and I think a lot about here in this whole discussion, is the patient perspective on some of these numbers so to speak, because there are people out there who can understand and are interested in the numbers. You might have a medical oncologist as a patient for that matter.

And it reminded me, too, that we’ve had a lot of posters over the years, our group, usually based on surveys we’ve done, but we’ve only had 1 oral that I ever had to present. It was actually at the ASCO GI Meeting, I don’t know, at least 10, maybe 15 years ago. And what we did, we tried to reproduce what was done in breast cancer. We went to people who’d gotten adjuvant therapy for colon cancer, and we said, “How much benefit would you have wanted to see for getting adjuvant chemotherapy?” And we gave them these charts that showed 10 out of 100, 3 out of 100, whatever. And in fact we saw that there was a significant cohort, about a third of these people, who would go through it again for a 1% difference in relapse-free survival, which kind of reminds everybody of the breast numbers.

Incidentally, no difference between men and women, no difference. And then the other thing was we saw a number of people, not as many, maybe 10, 15%, who wouldn’t take therapy again for a 10% absolute improvement in survival. So it really kind of showed the heterogeneity of how people look at these numbers.

Any thoughts about how you see people processing? Here you have a study that I don’t know whether you’d consider it negative or not, maybe it wasn’t as positive as we hoped, but it looks like it provided a lot of interesting information.

DR KOPETZ: Yeah, no. Absolutely. And I think it really does make that discussion very different.

For years I would prepare mentally when I went into a room to have a discussion about adjuvant therapy with Stage II, right, versus Stage III, and I would say okay, 30 minutes for the Stage II, 5 minutes for the Stage III, right? It’s kind of clear cut. Now it’s almost the inverse, right? The Stage II we know what we’re doing, ctDNA. If you’re positive, adjuvant. If not, not. And it’s the Stage IIIs now that we are having this much longer discussion as we get more of this information, and I think start to have to integrate this into what the patient’s wishes and desires are and what they hope to get out of adjuvant therapy. So great point.

DR LOVE: Also, it kind of gets into the downside of adjuvant therapy, in addition to the economic and convenience issues. To me, one big issue is neuropathy from oxaliplatin. And I’m curious, when you think back over the years, how much of a problem do you see neuropathy in the adjuvant setting?

DR KOPETZ: Yeah. It’s still an issue, especially for those patients that we are kind of committing to 6 months of adjuvant therapy, for example. I think it’s better now with our — where we’re using less; 3 months of adjuvant therapy for our lower-risk Stage III patients. But there’s still room to improve because we know that even some of what we think are classically high-risk patients really aren’t, and so we may still be overtreating a majority of patients.

DR LOVE: I would think that quite a few people might have baseline neuropathy already from type 2 diabetes. When you look at a patient like that, who has not — has some neuropathy to stat with, or people without neuropathy, roughly speaking, if you think about giving them adjuvant therapy for 6 months what’s the chance you’re going to have to hold or discontinue oxali?

DR KOPETZ: Oh, 90-plus percent of the time. So typically we tell patients that somewhere around dose number 8 to 10 is usually where we have to stop the oxaliplatin, acknowledging that there can be delayed neuropathy that can appear 6, 8 weeks after our last dose.

Prognostic and predictive role of ctDNA in the management of Stage III colon cancer treated with celecoxib: Findings from the CALGB (Alliance)/SWOG 80702 trial

Nowak JA et al. Prognostic and predictive role of ctDNA in stage III colon cancer treated with celecoxib: Findings from CALGB (Alliance)/SWOG 80702. Gastrointestinal Cancers Symposium 2025;Abstract LBA14.

DR LOVE: This is one of my favorite studies. This is from the GI symposium this year. I’ve always been fascinated by the lifestyle and the aspirin/celecoxib studies in colon cancer, and finally we get a kind of cool study here with the CALGB and SWOG. Can you talk about that?

DR KOPETZ: Yeah. This was a study that was initially done before the era of ctDNA asking — it was part of the IDEA consortium, 3 months versus 6 months, but there’s a second randomization that was there, and that was celecoxib or not. So this was celecoxib to continue daily for 3 years started with the adjuvant therapy.

So the report then said okay, now we have this tool, ctDNA, let’s go back and apply ctDNA retrospectively to this data. The initial study, we should remind ourselves, was negative, that in all comers celecoxib didn’t really add much. Maybe a little trend, a p-value of 0.1, but nothing there.

But the impressive thing was that when they went back and added in ctDNA — and now of course the ctDNA-negative patients did great; no benefit of anything beyond that. But, obviously, ctDNA-positive patients had a much worse outcome, but a substantially meaningful improvement in outcome, hazard ratio of 0.55 for the use of celecoxib in the positive patients, with a significant interaction.

And that showed up as an overall survival benefit there, as well, survival of 0.58. These are really meaningful differences in overall survival. Retrospective study, but still this is celecoxib, fairly benign, well tolerated for 3 years.

So this is really compelling because it’s at least suggesting ctDNA can be predictive of benefit of adjuvant therapy, so not just prognostic but actually tell us something that we could potentially do with that.

So there’s a lot of questions, like do we need to repeat this, or based on what we understand about the performance of ctDNA here is this something that should be changing practices now? I’m more in the latter. I’ve changed my practice on the basis of this, but it is something that is not yet in the guidelines but may find its way there in the near future.

DR LOVE: And are you using celecoxib or aspirin?

DR KOPETZ: So using celecoxib in this setting. There’s a second edition to the guidelines about the use of aspirin in patients that have a PIK3CA mutation, so that’s about maybe 1 in 5, 1 in 6 of our patients that have that, but for the remaining patients it’s really about using ctDNA to guide the use of celecoxib.

DR LOVE: And do you think the fact that they saw the benefit here was really an events thing, that they were able to get more events, and you can kind of see better what’s going on? Or do you think qualitatively the patients or the tumor is different?

DR KOPETZ: Yeah. I think you’re absolutely right. I mean, one way to think about this is if we had a perfect crystal ball, which is not ctDNA, but if we had a perfect crystal ball and knew who had micrometastatic disease and could separate them into those that did or didn’t, then of course we would say logically of course the benefit is only going to be in those patients that have — that have minimal residual disease and microscopic disease and not in those that were cured. So the closer that our assays get to the pure — the pure crystal ball I think the closer it will be to really defining a population that is only the ones that should be treated.

DR LOVE: Any comments on the mechanism involved? You were mentioning aspirin and the AKT pathway. What about celecoxib? Is it about the AKT pathway here?

DR KOPETZ: So there is some data that suggests that it could be through that kind of COX-2 mechanism that may impact KREB and feed-forward loop through the PI3 kinase pathway. The COX inhibitors also have some information on inflammation that — a signaling that may be relevant.

Aspirin has an additional ability to inhibit COX-1. It actually can decrease some of the tumor platelet aggregates in circulation. So I think there may be some overlapping biology, but they’re not identical to each other, and they may be working through different mechanisms.

Phase III ALTAIR study comparing trifluridine/tipiracil to placebo for patients with molecular residual disease after curative resection of colorectal cancer (CRC); a methylation-based, tissue-free ctDNA test

Bando H et al. A randomized, double-blind, phase III study comparing trifluridine/tipiracil (FTD/TPI) versus placebo in patients with molecular residual disease following curative resection of colorectal cancer (CRC): The ALTAIR study. Gastrointestinal Cancers Symposium 2025;Abstract LBA22.

Nakamura Y et al. Clinical validation of a methylation-based, tissue-free colorectal cancer test for the detection of molecular residual disease by ctDNA. ESMO GI 2025;Abstract 93P.

DR LOVE: So next paper. This is, again, from the ASCO GI meeting this year, the ALTAIR study.

DR KOPETZ: Yeah. So again, kudos to the leadership of our Japanese colleague, as part of the GALAXY study able to imbed a prospective randomized study. And this is the first prospective randomized study to look at a novel therapy, or at least not approved in adjuvant setting. And what they did is they took patients that were positive during surveillance and randomized them to receive trifluridine/tipiracil (TAS-102) or placebo because there is no standard of care in these patients after they’ve received adjuvant therapy and are part of surveillance.

The study showed some trends but unfortunately no real improvements in durable outcome. What we saw was a little bit of a delay. As you can imagine, using TAS-102 may delay recurrences for a bit, but unfortunately the curves then come back together. We’re not curing anybody by the addition of this.

So there was this trend, but unfortunately I would say kind of a negative study, but showing that we can do this study. And we now have other studies following in its footsteps. There’s an ADC study that is launching the AbbVie 400, the study to look at a cMET ADC. But we’re going to see more studies in this setting as well. So I think this was a good proof of principle, and it’s setting the stage for what I think is going to be a bit of a rush to develop novel therapies in MRD.

DR LOVE: Yeah. I guess, again, thinking about PSA in prostate cancer one big difference here is they’ve got endocrine therapy that everybody responds dramatically to, and we kind of struggle with colorectal cancer. So maybe if we have something more effective the benefit will be more apparent in some of these studies.

DR KOPETZ: Absolutely, and we hope that there’s things that we can do there. Immunotherapy, for example. Vaccines may be the perfect setting to deploy these in the MRD setting, where we know once the tumor’s bulky and widely metastatic they’re less effective.

DR LOVE: That’s a great point. Are there trials right now that you’re excited about using vaccines in colorectal cancer? Have you seen any benefits?

DR KOPETZ: There are, yes. Really interesting ones. I’m involved in one that is using a personalized mRNA vaccine, so it’s being run by BioNTech. That study’s completed enrollment. Waiting for readout of that. That is ctDNA-positive patients randomized to adjuvant with or without that personalized mRNA vaccine. That’s one of many. There’s also work being done with KRAS vaccines in this space, as well. So a lot of interest, for sure.

DR LOVE: Another assay, methylation-based tissue-free colorectal cancer test that was reported at ESMO GI. It sounds really interesting. Can you talk a little bit about what this is?

DR KOPETZ: Yeah. So a lot of the tests that are being done, the tumor-informed tests, are looking at DNA at mutations, but we also know that there can be epigenetic differences in the tumors that can help signal that there’s tumor in circulation and not blood.

So this was what we’d say is tumor naïve, so it’s not a tumor-informed or tissue-free assay, so it has some advantage of a little bit of faster turnaround time, and again, showing a very — very kind of impressive separation in these curves in this cohort in the surveillance window.

The questions remain about is there a 20% false positive, meaning patients that may be positive on this test and not have recurrent disease, so a little bit of a concern that there may be a bit of a higher false positive. But nevertheless a very strong assay here as well.

So I think the authors rightly said that the specificity was good, about 90%, and that this assay really could get turned around fairly quickly and give performance characteristics that were encouraging.

ctDNA with locally advanced mismatch repair-deficient/microsatellite instability-high solid tumors; real-world evidence regarding ctDNA with resected CRC

LaPelusa M et al. Long-term efficacy of pembrolizumab and the clinical utility of ctDNA in locally advanced dMMR/MSI-H solid tumors. Nat Commun 2025;16(1):4514. Abstract

Cohen SA et al. Real-world monitoring of ctDNA reliably predicts cancer recurrence and treatment efficacy in patients with resected stages I-III colon cancer. Ann Surg 2025;[Online ahead of print]. Abstract

DR KOPETZ: This is a study that we did looking at using immunotherapy in MSI-high primary tumors. So this was previously reported out that of course pembrolizumab, like many PD-1s, worked really well in the neoadjuvant setting, but also this idea to say can you look at an early timepoint and see whether or not a patient may be deriving benefit from this.

And so this was a report where — again, small numbers, so the caveat there, but we did kind of collect ctDNA and looked with this tumor-informed assay and were able to show a difference in both recurrence and a trend in overall survival, suggesting that an early timepoint in ctDNA may be able to really segregate those patients that are deriving the most benefit from the neoadjuvant therapy.

DR LOVE: What about ctDNA MRD analysis after 2 years of IO, at the point where maybe people are trying to decide whether to stop or not? Can you see a positivity there?

DR KOPETZ: So you can. It’s uncommon, but it is meaningful, right? So the idea to say at 2 years, when we normally would say well, let’s stop, and that’s what the studies have done, the majority of patients remain ctDNA negative at that point. But what we can see is that there are some patients that are still — that are still positive, fairly rare, at that later 2-year point.

DR LOVE: Any rough estimate of how often you see relapse after people are in CR for 2 years?

DR KOPETZ: It is single digits percentages, so the vast majority of patients. And it’s a big question that comes up, especially because many of these patients are not in a radiographic CR but still have nodules that are evident in the peritoneum or elsewhere, and these are really scar tissue. So this is where ctDNA can help, because radiographically you’ll say I still see something on the scans, and this can help reassure that no, it’s okay to stop at 2 years because there’s no evidence of disease remaining.

DR LOVE: That’s interesting. So some of these partial responses are actually complete responses?

DR KOPETZ: Are cures, right. Yeah. There was a series of about 15 patients that went to surgery for those nodules before we really understood them, and 14 out of the 15, even though we had centimeter or 2 lesions still remaining, those were scar tissue. That was just fibroblasts and no residual tumor cells.

DR LOVE: So interesting.

So this final paper I was curious about. We see so many real-world studies nowadays in oncology. A lot of times they’re very informative. I’m curious what your thoughts are about this paper.

DR KOPETZ: Yeah. So this is a cohort that was done that looked at over 6,000 timepoints, about 800 patients, and looking at some of these same features. Now, the difference here is this is real world. So these were providers that were ordering ctDNA and not part of a prospective study and so really gave us a sense of how often this was being utilized and the impact of it. And what we saw was that indeed in the real world, very similar to the prospective observational studies, that you get huge differences in them. Again, not randomized, but those that received adjuvant that were positive had better outcomes than those who didn’t post surgery, and that those that were MRD negative really no discernable difference in outcomes with adjuvant therapy. So reassuring that the real world is aligning with a lot of these cohorts previously have shown.

DR LOVE: I’m curious. In your clinical practice outside of clinical trial, and I don’t know how often you have patients who aren’t going into clinical trials because you have so many going on there, but if for some reason the patient’s not eligible, or just more of a theoretical thought, from your point of view do you believe that the current data supports or do you use yourself cell-free DNA in deciding about Stage II or Stage III disease?

DR KOPETZ: Right. For Stage II, absolutely. We’re utilizing it following the guidelines, saying positive it’s a high-risk factor and can be used for deciding adjuvant therapy. Stage III, again, we try to enroll on studies, but really it’s part of that conversation, when we talk about whether patients want that or not, the benefits of that therapy.

But we incorporate a lot of surveillance, a lot of testing after the fact, and thinking about how we can use that to titrate the intensity of our CT scans, obviously trying to get them on interventional studies if they’re positive. But even if patients can’t enroll on studies we know that diet, exercise, maybe aspirin, celecoxib perhaps, there’s a number of things that we can potentially do.

And I will say anecdotally nothing gets patients on a treadmill more sustained and more reliably than being ctDNA positive and having that conversation to say “alright, there’s something there, here’s our window.” The Canadian exercise study showed real benefits, let’s get going and see if exercise can help clear it.