Kasi Explores ctDNA and Other Advancements Reshaping GI Cancer Care

Pashtoon Kasi, MD, MS

The role of circulating tumor DNA (ctDNA) is rapidly evolving in the gastrointestinal (GI) cancer space and has the potential to revolutionize how experts diagnose, treat, and manage their patients.

ctDNA is being used to detect cancer at earlier stages, recognize the presence of minimal residual disease, monitor how a patient is responding to treatment, and identify genetic mutations in the tumor to personalize therapies. While it is still under development and its clinical utility is still being evaluated, its place in the field of oncology is emerging.

During a live event on Twitter/X, Pashtoon Kasi, MD, MS, of Weill Cornell Medicine and NewYork-Presbyterian Hospital in New York, New York, discussed the potential of ctDNA and other advancements that have the potential to improve the lives of patients with GI cancers which were covered at the 2024 Gastrointestinal Cancers Symposium.

Listen to the full event: HERE

Targeted Oncology: What were some of the most exciting themes or trends that you observed at the 2024 Gastrointestinal Cancers Symposium?

Kasi: Where advances were always every few years for patients, not just with colorectal or GI cancers but a lot of other tumor types, it is almost like at every conference, there is something tangible as an update. We had a bunch of updates for this year's ASCO GI event, and there were definitely themes regarding that, building upon the immunotherapy, subsets and treatments, [and] new targets like TIGIT. We had seen some data in hepatocellular cancer at ASCO last year, and we saw some of the data regarding that in esophageal cancer. There were themes regarding immunotherapy for the microsatellite instability [MSI]-high subset, looking at patients with specific biomarkers, and a lot of interesting work about moving things up into the neoadjuvant setting.

What were some of the biggest surprises or unexpected findings presented at the meeting?

Starting with upper GI cancers, the biggest advance in the last 5-plus years has been the addition of immunotherapy to the chemotherapy backbone. We saw results from a 5-year update on some of the initial trials with the addition of pembrolizumab [Keytruda] to chemotherapy. The fact that we are talking about 5-year survivors, where probably the word cure can be used in that scenario, with the addition of immunotherapy to chemotherapy for patients with esophageal cancers, was 1 thing that was of note.

We also saw some new targets and updates on trials. The SKYSCRAPER-08 [NCT04540211] trial looked at this novel anti-TIGIT [agent], which is another target alongside PD-1 for patients. This was with platinum-based chemotherapy in esophageal cancer. The median survival was about 15 and a half months vs 11 months. That was 1 target that seems like it is being investigated in many tumor types, including lung. Within gastric and esophageal, the subsets of biomarkers that are important [had] more educational pieces and also follow-up data on some of the trials that we have already heard about, the most talked about target being the Claudin18.2 in terms of another subset of patients who might benefit from a targeted therapy approach in addition to the chemotherapy. We are probably going to see the approval of the zolbetuximab, 1 of the drugs in that space, sometime this year.

In the subset of HER2-positivity, we had follow-up updates on the value of antibody drug conjugates and the sequencing of therapy as we have more options for these subsets. In the neuroendocrine tumor space, this is like the poster child for theranostics, this new field of therapy plus diagnostics withlutetium-177. In the GI space, our European colleagues have been using this for a long time and in the subset of neuroendocrine tumors in GI in the United States, the approval was not too long ago, but it was only for the low-grade, well-differentiated, the low Ki-67 subset of patients.

In this follow-up trial presented, this was inclusive of grade III. It also was moving it up a line of therapy and showed not just the efficacy, which was rewarding, but also to report on the small subset. With more long-term follow-up, there is anywhere from 2% to 4% that I have seen of results being reported out in terms of blood-related disorders, blood cancers, or early cancerous [myelodysplastic syndrome]-related bone marrow issues with these cellular radiation therapies that go through the entire body.

This brings up the question of the efficacy vs sequencing in this subset, but the fact that grade III patients can benefit, going back to the diagnostics idea that if you have a diagnostic scan that can tell you that the receptor or the target is present, then you could choose therapy accordingly, was a great example where despite the fact that these were grade III but well-differentiated, as long as they had the target expressed based on the scan, they were benefiting from it.

What were the biggest updates in the colorectal cancer space?

The theme was around circulating tumor DNA. There were trials, the BESPOKE study [NCT04264702] that we had the opportunity to present, as well as study from colleagues in Australia looking at the so-called DYNAMIC study [ACTRN12615000381583]. This was the first group that had presented the DYNAMIC-1 study, which was in colon cancer, showing that of the patients who had ctDNA-directed therapy, only half of them needed chemotherapy. With the same outcomes that resulted in stage II colon cancer population, the same group presented data in the setting of rectal cancer, patients who get the standard chemoradiation for the 6 weeks, and then they get their surgery. The question of, do we need to do more chemo[therapy] afterwards comes up all the time, and there are differing practices here. This group elegantly showed that maybe you could use ctDNA in the mix.

One thing I liked about the study was the fact that they weren't using ctDNA results in isolation. They had, in the treatment plan, 3 different cohorts. For somebody who is ctDNA-negative, the pathology report looked good, everything is negative, no treatment was recommended, and the patients were only getting observed. If somebody was ctDNA-positive, it was about escalation of therapy and getting chemotherapy. They also had a group in the middle, which is a great idea of combining clinical information with the ctDNA information, so that ctDNA is not in a separate room. It should be part of the clinical decision making. The pathology report is what we use in our day-to-day clinical practice. There's no reason why that report has to be hidden or isolated from the ctDNA results. Patients who were node-positive but ctDNA-negative were identified as a separate group where therapy was at the discretion of the investigator. It goes back to the question of keeping in mind, no matter how good the assay you have, you could have false negatives based on shedding or biology location of metastases.

GI Cancer and ctDNA : © SciePro - stock.adobe.com

The group also presented some really intriguing insights. When the ctDNA-negative patients recurred, it was often the lung metastases, and when the ctDNA-positive patients recurred, they often were liver metastases. It goes back to the idea of keeping biology in mind. ctDNA is here to stay, but there's no reason why it has to be a completely independent predictor of everything. I think the sensitivity, as well as the specificity, in my opinion, would increase even further if you could incorporate the clinical information that we know for our patients and the caregiver treating for that.

We also had the famous GALAXY study [NCT05986500] having updates. The cohort of data that's setting the benchmark is great, because then you could compare and contrast that with existing trials as to what to expect in terms of proportion of people who are positive in stage II vs III. The Japanese group had updates on that which would help inform the ongoing ctDNA-directed trials in the United States.

What research do you believe hold the most promise for future advancements in this space?

One subset is this MSI-high subset of patients that continues to amaze. This has been a story of serendipity of immunotherapy almost being not pursued anymore for colorectal cancer until this MSI-high subset was identified where immunotherapy has been durable. The word cure is the first thing that I started hearing in patients with MSI-high tumors, colorectal cancer initially, and that same story has continued to hold on to other tumor types.

There was the first readout of the combination immunotherapy from the CheckMate 8HW [NCT04008030] study that was looking at nivolumab [Opdivo]/ ipilimumab [Yervoy], which is a combination immunotherapy, CTLA4 with PD-1, vs chemotherapy. They also had an arm of immunotherapy that we do not have enough data on yet, but the hazard ratio in terms of the proportion of people who were cancer-free was striking at 72% vs 14% or so for chemotherapy. The hazard ratio was like 0.21, so almost a 79% reduction in the risk. It is a huge benefit, and you could see the plateauing of immunotherapy. Immunotherapy in the MSI-high subset of patients continues to amaze us, and it is holding true with these combination therapies even more.

Moving into ctDNA, can you discuss some of the specific challenges and opportunities in applying ctDNA technology to different tumor types?

It also brings to the table the fact that the broad term liquid biopsy or circulating tumor DNA is not just 1 entity. It started off with these advanced next-generation sequencing-based platforms, but quickly, in the last year or so, where the field has gone into is the minimal residual disease or the molecular residual disease, talking about any leftover molecular microscopic cancer that is not big enough to be seen, maybe a blood test can tell us so we can focus our energy and focus on sparing the toxicity to the folks who do not need it, who are likely already cured. We know for patients with stage II and stage III colorectal cancer that often, not knowing who is the one who's already cured, let's say with surgery alone, we often are treating more patients than we should be doing with chemotherapy, some of which can have long-term quality-of-life and [adverse] effects that can persist, like neuropathy. The idea of helping enrich the patients who might need, not just from standard chemotherapy, but maybe bringing into the equation some novel therapeutics, it is an exciting time with a lot of the clinical trials looking at ctDNA.

I think it is important to realize that you cannot just use ctDNA alone. Some of the trials that are only focusing on ctDNA-positive patients using a very specific technology could be falsely negative because of the assay itself, the biology issue, or the shedding issue. We also see ctDNA going down because of a treatment effect. In my opinion, ctDNA should help inform further and help allow more patients to participate in some of these clinical trials that are ongoing. It should keep in mind what's the standard-of-care and how to use the assays. We're learning it the hard way in some of these trials, where the only reason why somebody didn't start a trial is because of time lag or delay in getting the information.

We have to keep some of these practical considerations in mind as we interpret some of these assays. From the results of the trials, there is no doubt that ctDNA, in terms of colorectal cancer, is the strongest prognostic tool that we have. This has been compared across multiple platforms, across multiple studies, as the independent predictor of outcomes. We are seeing first hints of this being not just prognostic, but maybe predictive as well. We have to keep in mind as to how we are going to use this information and what is going to be the clinical utility.

Can you provide an overview of the BESPOKE?

BESPOKE was the first, largest, United States prospective study looking at a tumor-informed ctDNA. This was a study conducted at 133 sites from the country, and it kind of shows the enthusiasm in the field from the patients, caregivers, and providers alike. This study started in the middle of the pandemic and completed with over 1800 patients enrolled with stage II and III [colorectal cancer], and there was a cohort of patients with stage I, as well as stage IV cancer. This BESPOKE study at the conference focused on the stage II and III patients for which we had data in about 700 patients. The results go back to the same comment that ctDNA was the strongest prognostic predictor of outcomes.

The numbers were striking. If somebody was ctDNA-positive, if we had started with 100 people, only 30 of them would not have cancer by 2 years, meaning that 70% of the time the cancer came back. Whereas if somebody was ctDNA-negative, it's not 100%, but somebody could be cancer-free 92% of the time. The median survival was not reached compared with 16 months for the ctDNA-positive arm. We provided benchmarks of ctDNA positivity for patients using this tumor-informed platform. In stage II, which was about 6% to 7%, the expected rate of ctDNA-positivity in stage III, not surprisingly, was 3 times more at about 21%.

We also saw the same numbers and benefit in terms of prognosis in both stage II/III. The assay itself, more than the stage, was predictive. Going back to the predictive story, the benefit of chemotherapy was provocative, only in the ctDNA-positive patients. Their median survival almost tripled from 7 months to almost 18 months and if you look at the proportion of people at 2-year disease-free survival, it was about 12% vs 42%, so a big difference if you got chemotherapy vs didn't get chemotherapy if you were ctDNA-positive.

What was intriguing is that we are probably not yet ready to forgo chemotherapy in somebody who is stage III or somebody with high-risk stage II who warrants chemotherapy, even if the assay is negative. The fact that there was minimal or no difference with the median disease-free survival in the ctDNA-negative cohort whether they got chemotherapy or didn't, it was almost exactly the same. The confidence intervals were overlapping and anywhere from 90% to 93% of these patients were cancer-free at 2 years landmark where most of the times, if the cancers are to come back, they come back within a year or 2. It gives supporting data to forgo chemotherapy, at least in a trial setting, with the ongoing studies. Some of the trials that were started a year or 2 ago had some ethical considerations brought up by some oncologists, patients, and caregivers about whether it would be ethical or not to give chemotherapy to somebody who by the books should be getting chemotherapy. At least in the trial, it is randomized. It would be important to shout out to the CIRCULATE-US study [NCT05174169] and some of the ongoing studies. It is important and imperative to enroll in some of these trials.

We [also ask], what would we do if we catch the cancer early? Is ctDNA just increasing anxiety or is it something actionable as well? We were able to show that in 101 patients where there was cancer that came back, in about 40% of those, there was some oligometastasis-directed therapy, meaning that you could do some sort of surgery, freeze the tumor, do ablation, microwave, or radiation. There are 2 types of patients with metastatic cancer: 1 who has cancer in every single nook and part of the body vs somebody who has only a few spots in a few organs. The therapy for those is still curative in a multidisciplinary way, so it is important to keep that in mind.

On the BESPOKE part, we also looked at the patient's well-being in terms of how many of them felt that their anxiety was lower vs higher. Did they feel that they were receiving the right treatment? Overall, what we found was approximately 90% of the time, patients felt that they were receiving the right treatment when they got the results. They would continue to use the assay and they valued the additional information. We also did some other analyses. While somebody's fear may go up because their ctDNA is positive, so their cancer might come back, if [they are] not worried, that's a problem. It didn't necessarily lead to anxiety or depression, so it is important to keep in mind that ctDNA studies in the future should keep these lessons. It also brings the question of what about new therapies? How do we cure more patients, not just the MSI-high who have immunotherapy as a very viable option?

Please discuss the NEST-1 trial [NCT05571293].

The NEST-1 trial, a neoadjuvant study, piggybacked on the neoadjuvant paradigm shift that was happening for many cancers over the years. There is something about the cancer still being in your body, the lymph node still being there, the surgery not having yet perturbed the system and the immune system, that some sort of immunotherapy probably would be more conducive to be done in a neoadjuvant setting. We took that paradigm as this fertile soil to investigate new drugs and we particularly were interested in this drug called botensilimab [AGEN1181] with balstilimab [AGEN2034].

Not only did we see activity in an MSI-high subset of patients who had the MSI-high cancers, but we also saw significant shrinkage in kill of the cancer in a pattern, killing the cancer inside out, and also in a micrometastatic way. Assessment of ctDNA in these patients, piggybacking on everything from before with ctDNA-positive being not pleasant, and everything about ctDNA-negative being reassuring. The fact that we had more than 30 time points in the 12 patients that were on the study and everything is negative to date is provocative and has already led to the expansion of the study, where if this is the kill that we saw with just 1 dose and surgery just a few weeks apart, what will happen if we would wait a couple of months of treatment? The expansion for microsatellite stable [MSS] would be, as opposed to surgery at 4 weeks that was happening in our initial pilot study of NEST-1, the sequel to that would be to have at least 7 to 8 weeks of therapy before surgery happens to see if we could deepen the waterfall plot and responses.

For the MSI-high, we were pleasantly surprised. Some of our surgeons were saying that they are probably not going to operate on the next patient with MSI-high cancer who gets immunotherapy because they never find any cancer. It is getting even tougher for them to operate because of the scar tissue, because of the robust immune response that happens, which is a good problem. If we could cure people with drugs alone, we may or may not need surgery. There is a question of, can we even reduce reliance for surgery with drugs like botensilimab? I think that is the next step for the MSI-high subset. This MSI-high subset probably just needs to be separated from some of the existing studies because they are a different biology and have different outcomes, but it was rewarding to have some drugs or a combination of drugs. Granted, this is a neoadjuvant setting, but we have not seen anything this provocative in terms of data for patients who have mismatch repair-proficient or MSS or the so-called cold tumors, to have immunotherapy showing not just a signal, but deep and ongoing durable responses. There was a lot of interest and excitement from patients, caregivers, and experts alike. We are thankful to all the feedback that we got and are building upon the next steps.

How do you see ctDNA playing a role in routine clinical practice for GI cancer in the near future?

It is probably time for some guidelines to reconsider some of the recommendations that are there. We would continue to support ongoing clinical trials. There is no doubt about the fact that this should be an integral marker, as opposed to an exploratory marker in some of the ongoing studies. I think you could see from the real-world data, also patient quality-of-life data, the responses, the overall enthusiasm about it, and also keeping in mind what challenges we discussed. As clinicians, patients, and caregivers, the bottom line is, it is a shared decision making, so we cannot be dogmatic about this as a field.

Right now, what I'm seeing is polarized opinions, where there are institutions or experts who take pride in the fact that they have never done ctDNA as a test. There are institutions that are early adopters of this. It is not like ctDNA is going to solve everybody's problems. It is also not true that ctDNA is a waste of time. I think there is a role in value. It is another tool in our toolbox, with the clinical utility being demonstrated by the BESPOKE study, the GALAXY study, the DYNAMIC studies.

For example, for a patient with stage II low-risk cancer and their ctDNA is positive, is that somebody that we should talk about chemotherapy? I think that is a question that has come up in a lot of expert discussions, and most experts would say that is not unreasonable. For surveillance and closed monitoring in our subset, we were also able to show not just about new treatments and next chemotherapy. It is also about catching disease early where you could do something to it. I think the data regarding clinical utility continues to grow, and guidelines come with guidance. I think that would be even stronger in terms of utility in a judicious way, as opposed to having no guidance regarding what to do about ctDNA, and patients, caregivers, and providers doing it without guidance, which I would say is a bigger problem.

Are there any ongoing or upcoming clinical trials that you're particularly excited about?

I am excited about all these new drugs that are coming into the space. From an immunotherapy standpoint, I think there was a time where for cold tumors like MSS, people thought that the immunotherapies would not work. What we are learning is it may not be this particular commercially available immunotherapy, but there could be other targets. The number of targets continue to increase with the example of gastric cancer. Overall, I am excited about all the options.

From technology like ctDNA to novel classes of drugs to even things like cellular therapy making its way to solid tumors, as well as understanding that it is not just metastatic cancer, there is this neoadjuvant paradigm shift. There were also trials in progress with some of the KRAS vaccines, some of which were just published, and there were ongoing updates. I think it is overall an exciting time as an oncologist. From a patient and caregiver standpoint, these are important, promising, tangible things that are happening right now.

There is a line in the [National Comprehensive Cancer Network] guidelines that talks about the fact that the best management of a patient with cancer is in a clinical trial. If all of us can focus on helping, with not just enrollment, but even opening of clinical trials, coming up with novel clinical trial designs, that is where all the magic in terms of ideas is coming forward. It is unfortunate to see trials that never see the light of day because of whatever reason that often ends up being practical as opposed to scientific. I cannot overemphasize the importance of continuing to support clinical trials at every step of a patient's journey with cancer and also to remember that it's not necessarily your last-line options. It is a myth or misconception that you consider trials when everything else stops working. You would be surprised that there are trials now at every so-called line of therapy and even like our neoadjuvant trial before you did the first treatment, which was surgery. Overall, I think that would be important to move the needle forward.