Potential Clinical Implications of ctDNA as a Predictive Biomarker
George Ansstas, MD, and Firas B. Badin, MD, discuss the potential clinical implications of using circulating tumor DNA as a predictive biomarker in solid tumor patients treated with immunotherapy.
Firas Badin, MD: In summary, INSPIRE trial was a simple, straightforward study: 100 patients, different tumor types, progressive and conventional treatment, received immunotherapy with Keytruda, followed as you would usually follow in any clinic. In addition, we used personalized circulating tumor DNA and looked at outcomes, response rates, and survival and progression-free survival to see if that might correlate with what’s happening to the tumor DNA and blood.
I think there are a lot of interesting questions that we can go over from the [INSPIRE] study, multiple ones. I think the first question is, with any study, what is the key takeaway from this paper, George?
George Ansstas, MD: To me, it’s very important to my practice. Where I see most utilization of circulating tumor DNA, since I’m a melanoma physician, and I use tons of immune checkpoint inhibitors, is the phenomena of pseudoprogression. I feel this test now really relieves a lot of issues or problems in my mind. When I face these patients, if I have a baseline, and I follow the level of these patients, that could really help me in making my decision of whether I would proceed with my treatment or if I should change treatment on these patients. Second, actually, I’m learning more and more about circulating tumor DNA and the importance of reflecting on long-term outcomes. As you shared with me, Firas, here with patients who cleared their circulating tumor DNA, those patients had no relapses with almost a follow-up of 3 years. Such data make you feel that this biomarker can have a lot of value in our practice and can give us a lot of assurance in our decision-making. And lastly, I would also echo what we just said. We felt repeatedly with the data that the circulating tumor DNA can trump the imaging finding. Even if you have an assuring result on imaging, either if you have partial, complete, or stable disease, and you are assured with the results. But if you have an increase in circulating tumor DNA, you’re probably falsely assured that these patients are doing well. But in reality, those patients are destined to have progression soon after. That’s what I feel I’ve learned from the data out there from INSPIRE and some other collective data from Natera about the importance of integrating these things in my practice.
Firas Badin, MD: George, I’ve heard the argument from my colleagues, some of them are enthusiastic, some of them are not. Some of them say, “This is not the prime time yet for circulating tumor DNA assays.” But to me, really, this is the future of oncology. I think we’re taking our oncology clinic to a whole new level beyond conventional methods, beyond imaging, beyond radiology, beyond serology. Now, we have a way of really evaluating, very early, as we saw with the INSPIRE trial, as early as 6 weeks, whether the patients are going to draw benefit from the current treatment or not, and also predict their survival curves, that we can use this new method to help us clarify confusing clinical findings, like you mentioned, the pseudoprogressions, the pseudoresponders, the super responders. Sometimes, patients, if they clear their circulating tumor DNA, do we need to treat them the same way we do other patients? So there are a lot of very interesting questions that the study, at least theoretically, to me, has brought up. I personally might make some clinical changes in my practice, the way I approach certain patients, certain cases.
George Ansstas, MD: I agree, Firas, in all the points you raised here. Also, it’s food for thought for us who are in academia to build on this concept and these data to see how we can change our approach to many diseases. For example, I’m writing an investigator-initiated trial to look at the importance of clearing circulating tumor DNA as an early surrogate for outcomes in patients with a stage II melanoma. So we’re looking at patients with stage II for whom we have no standard of care treatment. Those patients traditionally are approached by observation, and serial scans, and examinations. We’re trying to approach these patients who have positive circulating tumor DNA, knowing that most of these patients, when they have this, they are destined to recur, with a treatment approach and to see how often we can clear their circulating tumor DNA. We are counting on the data you shared with me, that if we clear that circulating tumor DNA, probably these patients are destined to enjoy many years of relapse-free survival. I feel that what we raised here can have, definitely, implications either on day-to-day practices or even as a concept in many clinical trials.
Firas Badin, MD: George, based on these data, would you make any changes in the way you utilize circulating tumor DNA in your clinic?
George Ansstas, MD: I’m starting to use this more and more. I’ve been an adopter of this methodology probably for the past 6 months or so. I use it now more frequently to try to help myself in the clinic answering many of these questions that we raised, but also now, in patients for whom I’m really debating, in the adjuvant setting, whether I should treat or not treat, especially when we are in an iffy situation. I have patients who might be older and at higher risk for disease recurrence based on other traditional risk stratification. But the treatment could also hold some risk for them. So when I’m faced with many complex situations or complexity of findings, and I’m not really sure whether I should treat or not treat in the adjuvant setting, I feel that such a tool might be helpful for me. It might not be perfect, but it could help me in deciding whether those patients, if they have circulating tumor DNA positive in their blood, I feel those patients have a very high likelihood for relapse. Those who are negative, I might hold back a bit and repeat that test a few times. If it continues to be negative, I might be assured that these patients, I should not put them at risk for the treatment, given their other comorbidities. I feel like I’m playing on that notion of how important it is to know the status and translation for those patients for whom their disease might come sooner or later based on that biomarker.
Firas Badin, MD: Definitely. I think we have to be careful when we order personalized circulating tumor DNA assays. We need to have a question in mind. We should not just order this for everybody, for example. We need to have a question in mind. And also, we need to make sure that this assay will be able to address that question. You brought up a very good clinical scenario that I also personally use. We have more and more evidence that patients in the adjuvant setting, in GI [gastrointestinal] as well as in GU [genitourinary] and bladder tumors, that those patients, after surgery, if they still have that positive circulating tumor DNA, their outcome is not going to be the same with that treatment. And I use that argument sometimes for patients who are maybe reluctant to do adjuvant treatment, so I do use that assay. A very common place that I use it as well is if I’m ready to stop immunotherapy, for whatever reason, for toxicity, for patients’ personal wishes, they want a drug holiday, or maybe they reached their 2-year mark. Then this usually gives me confidence with my decision if I’m ready to stop treatment if the patient has cleared their circulating tumor DNA by the time I’m stopping the treatment. I don’t use 1 time point, I use more than 1, and I feel more confident that way. But if I don’t have a question in mind that I need this assay to help me address, and I do the assay and the personalized circulating tumor DNA comes back positive, it might really increase anxiety for me, as well as for my patient. So that’s why I’m not an advocate for ordering this. I find physicians either overestimate or underestimate circulating tumor DNA. I think if it is done in the appropriate clinical scenario, it’s extremely helpful for both patients as well as physicians. And maybe also, if you look at it economically, I think it can help as well in that way. If it guides our decision, treating when it’s necessary and not treating when it’s not working, I think we can save the system a lot of money.
George Ansstas, MD: I agree. I completely agree with your last point about how expensive the treatment is in oncology. And cutting costs on treatment that is not needed is a big advantage for the whole system now. Especially now, all our drugs are super expensive. If we want to utilize these drugs in the best manner, probably we should have a biomarker or surrogate for whether these drugs are working or not early in the treatment course.
Firas Badin, MD: Are there any interesting questions that this paper might raise for you, George, in your practice?
George Ansstas, MD: Definitely many interesting questions. There are some facts here, some speculation, Firas. Many of the points we raised, we might not have all the evidence behind them. We think this will translate in the future to be facts. There are also some facts. The interesting questions are the ones waiting on more trials to really prove what we think is right for many of the observations we just shared today.
Firas Badin, MD: I like the 6-week cutoff evaluation. A lot of times, in my practice, I don’t know about you, George, but we use 3 months to evaluate response and maybe even sometimes longer. But now, maybe we can evaluate sooner. Maybe we don’t need to keep patients on 3, 4 months of treatment if it’s not working. Maybe we can use 6 weeks as a cutoff. And with the INSPIRE data, 98% is pretty damn good. Ninety-eight percent. If you don’t decrease your circulating tumor DNA by 6 weeks, 98% of the patients, they’re not going to respond to treatment. So I find that really very interesting. And hopefully, we can build on it for future studies. Also, I like when they looked at the circulating tumor DNA, as well as the imaging findings, and seeing circulating tumor DNA is more accurate than when we always relied on CT scans, or PET [positron emission tomography] scans, or response by imaging. Even if you have any kind of response or stability in imaging, if you see that circulating tumor DNA is not going the right way, it’s not going down, probably with those patients, their outcomes are not as good as other patients.
George Ansstas, MD: Very true. To echo what you said about the trial, this is the same concept I’m utilizing in the trial in stage II melanoma, to look at that early end point, the clearance of circulating tumor DNA before cycle 3 of immunotherapy, within that time frame, like 6 weeks. If we prove to be right repeatedly and have the same observation, probably this observation would be practice-changing in the future.
Firas Badin, MD: Are you doing any studies with circulating tumor DNA in your practice? What’s the next step? How do you take it from here?
George Ansstas, MD: We’re opening the trial in phase 2 also for stage II melanoma. For patients who have positive circulating tumor DNA, the plan is to have a treatment arm only for these patients and then to look at the clearance of circulating tumor DNA before cycle 3 and to look at the outcome of these patients in comparison to those who have persistent circulating tumor DNA. So we’re hoping to see some positive finding on that one because if we see early clearance, this could be a very good predictor of long-term outcome of these patients. That’s what we’re really trying to prove in a prospective manner in patients with melanoma. Also, we are trying to be part of the BESPOKE clinical trial as well. We’re trying also to accrue more patients from lung, colorectal cancer, and melanoma, and look at these patients in a prospective manner to see their outcome, how it plays when these patients are treated with immunotherapy, and to correlate that with their circulating tumor DNA trend. Lastly, we are part of the early adopter program for patients with stage II melanoma. This is a lung treatment arm, but we’re really capturing the natural course of these patients. As I said earlier, those patients are being just observed by clinical exams plus/minus imaging. We’re trying to see, if we monitor these patients with circulating tumor DNA, how their course will play out in correlation to positive versus negative tests on these patients because, again, it could be in the future a good biomarker to dissect these patients, whether they will need treatment or not. Those are what we’re involved in, Firas, in terms of trials and trying to answer some of these questions.
Firas Badin, MD: To me, George, I think given the innovative technology that those companies use with detecting a minute amount, very sensitive, very specific when it’s a personalized assay, there are different venues to utilize it in an oncology clinic. Minimal residual disease is definitely one of them; like you mentioned, George, in adjuvant setting trials; melanoma; really any other tumor type. That’s the beautiful thing about this. It’s really not disease specific. It can be utilized in any tumor type that we have. Now, we know certain tumor types might secrete more DNA than others. But I think as the technology improves, even a very small amount has been able to be detected. So any time we have a question about adjuvant treatment, I think this assay might be helpful. Any time we have confusion because of unclear imaging or serology, I think this assay might clarify it for us. And then any time we are wondering about duration of treatment, and is 1 year what we need, or 2 years, or 3 months maybe for super responders, I think this assay might be very helpful. This is just the beginning, but I think the future looks quite promising.
George Ansstas, MD: I completely agree, Firas, with all these notions. And stopping treatment early, not until disease progression or an acceptable toxicity, probably is a big area of improvement and an unmet need for many of our patients. I see what you just said could be clinically translated into many of our patients after they get a sufficient amount of treatment, as you said, like a year or two, to feel more comfortable about stopping immunotherapy treatment based on such a biomarker.
Firas Badin, MD: Well, George, thank you so much for your time today. And thanks to CancerNetwork® for this opportunity for both of us to be able to talk about a very interesting topic.
George Ansstas, MD: Thank you so much.
Firas Badin, MD: In summary, the INSPIRE trial enrolled about 100 patients with different tumor types who progressed under a standard treatment. They were either not candidates for further treatment, or they did not have any other treatment options. They received Keytruda [pembrolizumab], 200-mg IV [intravenously], every 3 weeks. In addition to conventional follow-up methods to evaluate response, the study also used a personalized circulating tumor DNA assay that was captured at baseline, then every 3 weeks, up to 6 months. They also looked at outcomes, which were response rate, progression-free survival, as well as overall survival, to see whether that did correlate with circulating tumor DNA or not. They looked at the correlation also between circulating tumor DNA and conventional methods to evaluate response. As we saw with the study results, if the circulating tumor DNA did not decrease by as early as 6 weeks, those patients usually are not going to respond; 98% of them will not respond. Also, there was a significant improvement in survival, as well as overall and progression-free survival, in patients who had a decrease in their circulating tumor DNA compared to patients who had an increase in their circulating tumor DNA compared to the median.
Transcript edited for clarity.
