I’ve incorporated ctDNA [circulating tumor DNA] into my practice for the last 2 years. Some of it is a part of clinical trial work, other [parts of it are] standard of care. I’ve found it to be 1 of the most powerful pieces of information that I have for patients. There’s no question that, for your stage II patients and perhaps even some of your low-risk stage III patients, ctDNA can be very helpful. For patients with IIA and IIB colon cancer, I tend not to treat most of these patients with chemotherapy based on the available data, but finding the patient circulating tumor DNA negative gives the patient and me reassurance that we’re probably safe not treating. On the other hand, if they’re ctDNA positive, I treat these patients understanding that the question is subject to the trials that we’ve discussed. I’d enroll these patients in clinical trials and CIRCULATE-US when they become available. Absent that, I’d give the stage II ctDNA positive population chemotherapy. For some of your IIIA patients, you could consider the same. That’s a very low-risk population that probably doesn’t need even 3 months of chemotherapy. If they’re ctDNA negative, is that a population that can watch? From an informed decision-making standpoint, I found that to be helpful for patients to see, and it gives reassurances.

The other place that I found a lot of interest in ctDNA is in the stage IV NED [no evidence of disease] population. For patients with liver and lung metastases who we’ve resected, we’re making the decision whether to get chemotherapy. The randomized clinical trials tell us that we don’t benefit these patients with chemotherapy in practice, but I find that most oncologists have a hard time not giving a patient with metastatic disease resected chemotherapy; it doesn’t feel right. The surgeons don’t like us not giving chemotherapy to this population, but if I find them ctDNA negative, then the patient and the surgeons seem to calm down a bit.

Alternatively, if they’re positive, that’s a population that we should probably treat. I could be wrong, it may be the reverse of that. Maybe we should be aggressively treating ctDNA negative and not treating ctDNA positive. But until other information comes out, it gives me assurance to not give chemotherapy when the data support not doing so. I try to isolate high-risk patients who would benefit most from chemotherapy.

In our tumor boards, we’re constantly bringing up ctDNA results and recommending ctDNA testing to patients. When you consider the overall strategy and technology, this should be a tumor-agnostic strategy. There’s no reason that this technology shouldn’t work in pancreatic cancer, lung cancer, breast cancer, etc. In the GI [gastrointestinal] world, I’m excited to start looking at it in our patients with pancreatic cancer. In the colon world, we’ve got a lot of good data the most powerful prognostic indicator postoperatively, in terms of ctDNA positivity or negativity, is much better than CEA [carcinoembryonic antigen]. This is something I encourage people to start incorporating into their practice. As CIRCULATE-US becomes available, if you have the apparatus to run clinical trials, this is 1 of our most critical trials. It seeks an answer: can we de-escalate therapy in the negatives and escalate therapy in the positives and improve outcomes for these patients?

Videos

Advice to GI oncologists in the community on using ctDNA testing to detect minimal residual disease and inform treatment decisions in colorectal cancer.

Incorporating ctDNA Into Routine Practice: Clinical Pearls

 In terms of other abstracts presented at ASCO GI [American Society of Clinical Oncology Gastrointestinal Cancers Symposium], we have many variations on the themes I’ve discussed. There are other prospective cohort studies looking at ctDNA [circulating tumor DNA] and trying to refine the process. There’s a study using the Guardant Reveal assay, which is nontumor informed. It’s essentially an off-the-shelf test using a defined set of DNA markers as well as methylation status, trying to answer some of the questions that Natera has been looking at in the multitude of their studies. The advantage of something like Guardant Reveal is that it technically doesn’t require any tissue. There are situations where the tissue is hard to obtain, or the patient received neoadjuvant therapy for rectal cancer and the tissue has been obliterated; that assay may prove important.

There’s a lot of interest in looking at the dynamics of ctDNA change. Does it matter whether ctDNA at 2 time points stays stable or if it doubles or triples? For the clinicians out there, this is thinking about PSA [prostate-specific antigen] velocity, a concept many are familiar with. The idea is to look at quantitative change in ctDNA. There are preliminary data suggesting that quantitative change in ctDNA is prognostic of relapse. How can be used? Let’s say a stage IV NED [no evidence of disease] patient has positive ctDNA, and you and the patient aren’t excited about chemotherapy. You check another ctDNA 3 months later, and it’s doubled or tripled. Then you can confidently say that this cancer will return quickly. Whether my therapy will prove beneficial there, this must be studied. Similarly, patients with ctDNA who remain positive but stable are a population that you can watch more closely over time with serial imaging, and you don’t need to act until you find something.

There was another work looking at ctDNA published by our group at the University of Chicago looking at appendiceal cancer. We have a large volume of this peritoneal-only disease that displayed very good test characteristics for ctDNA positivity in patients with known peritoneal disease. This is 1 area that we weren’t sure how well we’d be able to pick up. It’s not as good as you would see with solid tumors in liver or colon cancers, but we’re finding that it’s useful in peritoneal disease. We’re beginning to see that the quantitative level of ctDNA helps us predict patients who have peritoneal-only disease vs patients who are at risk for visceral metastases. Those are all the areas: expanding it to some of these more unusual settings; looking at the nontumor-informed assays, which may have some advantages; and seeing whether the test characteristics hold up as well as those seen with Natera’s tumor-informed assay. Quantitative ctDNA is an important marker by itself and as a prognostic marker in these patients.

Abstracts presented during ASCO GI 2022 and other initiatives in clinical trials, assessing treatment strategies for patients with colorectal cancer based on ctDNA negativity or positivity.

Investigating the Role of ctDNA in Colorectal Cancer: Next Steps

 CIRCULATE-Japan and CIRCULATE-US have the same name because it is part of the IDEA collaboration, which was the worldwide successful endeavor to look at 3 vs 6 months of chemotherapy in patients with stage III colon cancer. That has changed the standard of care for many of our patients moving to 3 months of therapy. This is a coordinated effort and asks the questions, “Do we need to treat patients who are ctDNA [circulating tumor DNA] negative with chemotherapy? Can we escalate therapy for patients who are ctDNA positive?” The VEGA study looks at whether we can de-escalate, and CIRCULATE-US has a similar arm. If you look at VEGA vs CIRCULATE-US, the CIRCULATE-US study looks at patients with low-risk stage III cancer—stage IIIA and IIIB. If they’re ctDNA negative after therapy, they’re randomized to CAPOX [capecitabine, oxaliplatin] or FOLFOX [5-fluorouracil, leucovorin, oxaliplatin]; the other arm is surveillance. Essentially, it’s 3 months of chemotherapy with CAPOX [capecitabine, oxaliplatin] or 6 months with FOLFOX, or observation. It looks at disease-free survival as a primary end point.

Cohort A is what I just described. The second part of that study is called cohort B, which is like the ALTAIR study in CIRCULATE-Japan. It looks at patients with stage II or IIIC cancer who are ctDNA positive; that’s the first group. The other group is patients who are in the surveillance arm of cohort A who turn ctDNA positive; that’s cohort B. It’s either de novo stage IIIC and II colon cancer that’s ctDNA positive postoperatively or stage IIIA and IIIB randomized to the surveillance arm in cohort A who subsequently become positive. Patients in cohort B will be randomized to 6 months of CAPOX [capecitabine, oxaliplatin], FOLFOX [5-fluorouracil, leucovorin, oxaliplatin], or an escalation arm of FOLFOXIRI [5-fluorouracil, leucovorin, oxaliplatin, irinotecan]. The United States is looking at a much more aggressive approach than CIRCULATE-Japan in terms of our escalation. They’re using Lonsurf [trifluridine, tipiracil]. We’re inquiring about incorporating irinotecan in this high-risk population and if that will provide us a disease-free survival advantage. These are the kind of studies we need to be doing. CIRCULATE-US should be opening in early 2022, it’s sponsored by NRG and SWOG, so it should be available to many centers around the country. This is probably 1 of the most important trials we’ve done in adjuvant colon cancer in a while because it can change the standard of care. If both trials’ hypotheses are proven, then we’ll change the standard of care. We’ll stop giving chemotherapy to a large percentage of stage IIIA and IIIB patients. We may be able to find a way to rescue patients who are ctDNA positive with a more aggressive chemotherapy backbone.

An overview of the design of the CIRCULATE-US study, set to further explore the potential of ctDNA testing in colorectal cancer.

Objectives of the CIRCULATE-US Study in Colorectal Cancer

 I want to discuss the CIRCULATE-Japan program. It’s a somewhat complex design, so I’ll briefly describe the overall design of the trial and then focus on the GALAXY study. CIRCULATE-Japan is a 3-arm study of patients with stage II and III colon cancer or rectal cancer. They eventually will be included in the GALAXY study for patients with resectable M1 disease, so patients with liver or lung metastases. GALAXY is a prospective cohort study. It checks preoperative ctDNA [circulating tumor DNA] and postoperative ctDNA every 3 months for a 2-year period and looks at what happens to those patients. There’s no defined therapy in GALAXY. It’s simply looking at a cohort with usual care according to the physicians in Japan. They have a trial that they’re launching called the VEGA trial, which looks at patients who are ctDNA negative postoperatively and randomizes them to a control arm that will or will not get standard chemotherapy. In their trial, it will be capecitabine-oxaliplatin for 3 months vs the experimental arm, which is no chemotherapy. This is 1 of the more exciting studies and elements that we have. This will hopefully answer the question in a prospective fashion: can we deescalate therapy in ctDNA-negative patients?

This the third part of the trial. GALAXY is the cohort study, VEGA is the de-escalation trial, and the ALTAIR trial looks at patients who received adjuvant chemotherapy and are post-op ctDNA positive, or patients who were in the experimental arm of VEGA who didn’t get chemotherapy but turned positive at some point. They’re looking at giving 6 cycles of Lonsurf [trifluridine, tipiracil] therapy vs placebo, asking the question: if you’ve been appropriately treated with standard chemotherapy or weren’t treated but turned positive, can we use an alternative chemotherapy like Lonsurf and change the outcome in those patients? One can hypothesize that there’s very little we can do for ctDNA-positive patients. It’s not an unreasonable hypothesis that if you are ctDNA positive, your biology has declared itself, and no matter what we do your cancer will return. This is beginning to ask that question.

Focusing on the GALAXY study data in more detail and the data that we saw at ASCO GI [American Society of Clinical Oncology Gastrointestinal Cancers Symposium] in 2022, we’re looking at the positivity rate of patients preoperatively. Patients with stage I cancer had an 80% detection rate, and there were 96% detection rates in stages II and III. Postoperatively, patients who were stage II and III had about a 6% ctDNA positivity rate, compared with 25% to 30% in stage III and IV patients and stage IV–NED [no evidence of disease] patients who had their tumors resected.

The GALAXY study confirmed what we’ve seen from the [Thomas] Reinert trial data, from the extension, and the [Tenna] Hendriksen trial data: that ctDNA predicts relapse on average about 8 months before you see it on a CT scan. Whether you think that’s a good or bad thing depends on your perspective, but there’s no question that it’s a more sensitive and specific marker than anything else we have, and it’s certainly better than CTs in detecting disease. Whether that lead time is helpful, whether we’re just telling patients something that happens but with no strategy to treat them, from recent trials hopefully we’ll have a strategy to do something early before they develop macro metastatic disease.

Implications of the GALAXY study in CIRCULATE-Japan, in terms of utilizing ctDNA to guide decisions for adjuvant chemotherapy in colorectal cancer.

Updated Results of CIRCULATE-Japan in Colorectal Cancer

 The [Tenna] Henriksen study published in 

] in 2021 is a sample of 265 patients with stage II and III colon cancer who underwent resection. It looked at the predictive and prognostic value of a positive ctDNA [circulating tumor DNA] postoperatively. First, the overall data: about 9% of patients had ctDNA positivity postoperatively; 75% of those patients relapsed. The 25% who didn’t relapse received adjuvant chemotherapy. That’s important because, from a hypothesis-generation standpoint, it begins to say that there are patients who are ctDNA positive, and if we give them chemotherapy, we can convert them to ctDNA negative and perhaps decrease the chance of a relapse.

All patients who did not relapse got adjuvant chemotherapy, but many who got adjuvant chemotherapy did relapse, so it’s not perfect. Adjuvant chemotherapy does not reduce the risk to 0, but some portion of patients benefit from it. When you compare that with the ctDNA-negative patients, 13.6% of them relapsed, but there’s about a 75% relapse in the ctDNA-positive patients. Those are very large differences between ctDNA positive vs negative. If you look at ctDNA positivity vs ctDNA consistent negativity at any point over a 2-year period, patients who are ctDNA positive have a 36 times higher risk of relapse compared with those who remain negative. There’s no question that we’ve found 1 of our most powerful prognostic factors in predicting colon cancer relapse. We’ve identified the best way to detect which patient population will relapse. We don’t know if we can benefit the patients who are ctDNA positive by giving them more aggressive chemotherapy. Alternatively, can we choose not to treat patients who are ctDNA negative? The predictive nature of the assay remains to be seen, but I’m very excited to say that there are worldwide efforts to try to answer that question.

Key takeaways of a recent article highlighting the use of ctDNA analyses for the assessment of recurrence risk, benefit of adjuvant therapy, and early relapse detection after treatment in colorectal cancer patients.


Earlier Detection of Relapse in Colorectal Cancer

 Clinically, how would we use ctDNA [circulating tumor DNA] testing? What do the data show? In most studies in the stage II and III colon cancer setting, the Signatera assay has Medicare approval. Why did they get approval for it? There have been several studies. The [Thomas] Reinert study, published in 

[in 2019],is probably the first that we saw. At ASCO GI [American Society of Clinical Oncology Gastrointestinal Cancers Symposium] in 2022, we saw additional data from that group, as well as data that came out of the CIRCULATE-Japan group.

If you look at a onetime circulating tumor DNA after resection, if you’re ctDNA positive, the data from Reinert and colleagues show that 75% of those patients will have a relapse. When you look at the data further at the patients who did not relapse, almost all those patients received adjuvant chemotherapy. As you extend it further and look at ctDNA positivity at any point within a 2-year period—every 3 months ctDNA testing, if anyone is positive in that period— 97% to 98% of those patients will recur. Now it’s 9% to 10%, depending on stage of patients postoperative, stage II and III, who will have positive ctDNA. For those patients, the risk of recurrence is exceedingly high.

If you look at it from a relapse standpoint, 4 to 6 weeks after surgery, patients who are positive have a chance of recurrence that’s 11 times higher than those who are ctDNA negative. When you extend the data out—patients who are negative for over 2 years vs positive at any point over the 2 years—the risk of relapse for the positive patients is 36 times higher. We have a test that has good test characteristics in terms of sensitivity and specificity, and positive and negative predictive values are in the 97%-to-98% range. If you look at patients who are ctDNA negative for 2 years, those patients have less than a 3% chance of relapsing. You can imagine how this may impact our surveillance as well as our chemotherapy decision-making.

Look at the data that came out of the GALAXY study—this is a CIRCULATE-Japan study. It’s a very large multiarm study. It’s essentially an observational cohort of patients who have ctDNA. In that population you see similar data. They did preoperative and postoperative testing. In the preoperative setting, when you look at patients by stage of disease, of patients who have stage I colon cancer, 80% have ctDNA. When you look at stages II and III, 95% to 96% of these patients have ctDNA. Compare that with CEA [carcinoembryonic antigen] stage II colon cancer. CEA is positive in about 35% to 37% of those cases. Stage III cancer CEA is positive in about 48% of those cases, compared with 96% in patients who have circulating tumor DNA. There’s no question that circulating tumor DNA is picking up cancer.

When we look at the postsurgical setting, surprisingly, 6% of patients are positive for ctDNA for stage I disease, which we would never consider giving chemotherapy to. For stage III patients, as many as 25% are positive postoperatively. We have a test that detects disease when it’s there at a very high rate. It’s detecting disease postoperatively when we don’t see disease by imaging in a high percentage of our lower-risk cancer patients and in a quarter of our stage III patients. We’re beginning to define and see a population that we may want to be more aggressive on or, alternatively, give less aggressive therapy to.

How ctDNA testing can be used in the preoperative and postoperative settings to help oncologists make informed treatment decisions.

Preoperative and Postoperative ctDNA Testing in Colorectal Cancer

 The field of ctDNA [circulating tumor DNA] is rapidly evolving. As oncologists and clinicians, we’re beginning to figure out how to incorporate it in. We’ll discuss some data and trials, but it’s important that we ensure that we have a common understanding of what we’re talking about. ctDNA analysis can come in several forms. One is a tumor-informed ctDNA, which has been pioneered by Natera. Many of the abstracts that we’ll discuss today use the Natera’s Signatera assay. This looks at the patient’s tumor, creating a proprietary set of screens for ctDNA in the blood and developing a quantitative score of the amount of ctDNA circulating in the blood. The other basic approach is a nonpatient informed assay. For example, the Guardant Reveal assay looked at a predefined set of markers and methylation status to give a positive or a negative, not a quantitative level. Those are the 2 platforms available. Neither will give you mutations, because with Signatera or Guardant Reveal, you don’t receive mutations. It’s not a typical Guardant assay where you receive a specific tumor mutation data; you’re looking at, is there or isn’t there circulating tumor? In the case of the Signatera assay, what is the quantitative level?

Today we’ll discuss the stage II and III colon cancer setting. There are many patients who we overtreat, meaning patients who are unlikely to have their cancer return and are subjected to 3 or 6 months of chemotherapy, so we can better define the high-risk population and perhaps leave those who are at lower risk alone. The other interesting area that these assays are beginning to study are areas of minimal residual disease in the stage IV colorectal cancer setting. Another area subject to controversy is what we do with patients who have liver or lung metastases that we’ve resected. Should we give them additional chemotherapy after resection? If you look at the evidence-based data, there are no data that suggests any benefit to adjuvant chemotherapy in this population. In fact, some data from Japan suggest we may be doing harm by giving chemotherapy to this population. We can imagine looking at ctDNA after resection, identifying higher-risk patients and seeing whether chemotherapy benefits that population and if it’s advantageous.

The final area we’re beginning to see data on ctDNA is immunotherapy. For example, the KEYNOTE-177 trial, which was pembrolizumab vs chemotherapy in patients with frontline metastatic colorectal cancer who were mismatch-repair deficient. Pembrolizumab had a significant advantage over chemotherapy any way you looked at it, but a third of the patients rapidly progressed on immunotherapy. There are data in immunotherapy with Signatera asking if we can look at a 6-week Signatera for patients with ctDNA that is unstable or decreasing. Increasing will likely be that third of patients who need more aggressive therapy. It allows us to initiate chemotherapy earlier in that patient population. The individuals who respond to immunotherapy continue it for the KEYNOTE -177 data. That’s the landscape of areas. We can discuss what the trials will try to answer. I know there are some burning questions among those of us in the clinical oncology and GI [gastrointestinal] oncology space.

An overview of the ctDNA testing platforms and the potential clinical applications of ctDNA testing across the colorectal cancer disease continuum.

Blase Polite, MD

Articles