Role of ctDNA MRD Testing in Immunotherapy Selection for NSCLC: Precision Oncology Breakthrough vs Early Clinical Evidence (Round 3)

To hear more arguments related to testing in NSCLC, see the other videos in this series.

Dr. Garon: So, I think that one of the big questions obviously these days is many of our patients are getting IO therapies. And nearly all patients outside of the driver mutation, positive patients, do receive immuno-oncology therapy as part of their routine care. And so, here again, what you're looking at are outcomes. We actually saw a snippet of this slide earlier. And now, you're seeing sort of—it's sliced and diced in multiple ways. But, again, looking at outcomes in patients—this is immunotherapy-treated patients, whether they were MRD positive or not. And I always think one of the things that's interesting, you start looking at the hazard ratios here. And, in general, for those of you who don't conduct a lot of clinical studies, when we are looking at drug studies, we're trying to show a hazard ratio of, like, 0.7, for instance. And that, if you look at it, the other way of looking at it is 1 over 0.7 for what the benefit or risk is. So, if you have a hazard ratio of 0.7, it's nearly 50% better over the entirety of the curve to have gotten this new intervention. When you start looking at the hazard ratios on something like this, we're looking at numbers in the double digits often. So, it's 10 times worse over the entirety of the curve if you have—that's what a hazard ratio of 10 would be if you were to have positivity for MRD in this case. And so, I think that obviously this is a profound difference in how one would evaluate how people are doing. So, one might ask what is really the clinical relevance of this, right? What are the decisions that can be made? And one of them, of course, is it has to do with whether or not we should continue therapy. So, this is obviously a hot topic in management today. If somebody gets neoadjuvant therapy, should they receive adjuvant therapy as well? Should they receive additional therapy? And we have one marker that oftentimes will tell us that we don't move forward with adjuvant therapy. And that is patients who have a complete pathologic response oftentimes we will not give additional immunotherapy. Interestingly, this topic is quite different in the US as opposed to Europe or Canada where there's socialized medicine. In the US, we tend not to have a big impediment to offering adjuvant therapy in places where the government is paying for the health care. They obviously have a very legitimate right to say, where's the data to give many months of additional immunotherapy without a survival advantage? And we've generally looked at these as just sort of two main groupings, either your, you know, how do the people do if they got adjuvant versus not? This is an analysis where Bristol-Myers Squibb with nivolumab was able to look—they had one study where everybody got only neoadjuvant chemoimmunotherapy. They had one where people got neoadjuvant chemoimmunotherapy followed by nivolumab. And they were able to compare the outcomes. So, if you look in aggregate, the outcomes are better for the adjuvants. But one question is, would you be able to pick out a group of people that really don't need additional adjuvant therapy and forego months of coming into the clinic for the adjuvant therapy and a very tremendous cost? And I think that's one very significant potential benefit. And we also are starting to see clinical trials where, in fact, we're trying to take the other group. So, it's obviously reassuring to be ctDNA negative. And maybe that's a group that we could argue, maybe the adjuvant immunotherapy is not needed, or we could stop earlier, for instance, in that group of patients. On the other hand, in the people who are MRD positive after surgery, things aren't going to go well. And I think if they've already gotten, you know, three, four cycles of chemoimmunotherapy, the likelihood that what they really needed was five or six cycles or a dozen more seems very unlikely. And so, there are many trials that are currently underway looking at giving additional therapy in that setting. So now, I'll hand this over.  

Dr. Raez: I agree with Dr. Garon that, for example, this is a very practical point that if we are giving chemoimmuno as it’s becoming a standard of care now, it will be perfect after the surgery to do the MRD. And if you are negative, we don't use more adjuvant therapy because the problem is before the surgery everybody needs the chemoimmuno because the tumor is there. But, after the surgery, you have a bunch of people that are cured. And now, we are giving one year of immunotherapy without knowing if the patient needs or not. So, that's why the MRD will be perfect after the surgery. If you do the MRD, it’s negative, you don't get treatment. If you do the MRD, it’s positive, you do get treatment. The problem is that the studies that we have now do not support the evidence yet, so we still have to use other factors to decide. And this is a good example. For example, this is Impower010. This is the study that we give atezolizumab after surgery. You look, in your left, the ctDNA is positive. It makes sense. You know, without therapy, you are in red, your disease-free survival. With therapy, you go to blue. Makes sense. But this is what I was telling you, the treatments are bad. Okay, the treatments are bad because you are positive. So, your survival without therapy is the red. We’re supposed to give you a therapy that cures you, you know? And we, unfortunately, with the technology that we have for immunotherapy only elevates the curve. But still, most of the people are going to recur. So, what does it tell you? One of the challenges is don't blame the MRD. We also need better therapies. And this one is even more—this is the ctDNA is negative. This is the people that are supposed to be cured. But they are not getting cured, you see? At four years, half of the patients are going to fail. But this is the group that is ctDNA negative. So, our MRD is supposed to predict, and it's supposed to be flat, you know? So, that's, of course, you know, they still get therapy. We gave therapy to everybody, because the therapy for this study was not based on MRD. But this is the challenge in why we don't use MRD yet, because we need to make it better. If MRD is negative, you're supposed to be cured. If MRD is positive, you're supposed to recur, and we are supposed to find a treatment that makes a difference. And I'm not negative. I am very glad that we discovered a test. So, you know, we save a lot of people now. We even cure people with that test. But we don't cure everybody. Our goal, our bar is 100%. And basically, the rest of the studies, I will show you the same thing. These are realist studies. This is AEGEAN. This is the study from AstraZeneca that shows that the perioperative is, you know, you give neoadjuvant, surgery, you give adjuvant. And, of course, in all of these studies with the ctDNA, look at this challenge. We were able only to construct MRD in 38% of the patients. That's the challenge. There is not enough tissue. Why you put in the bar in the MRD when it's not their fault? There is not enough tissue. And you can see the same problem here, similar, no? People with ctDNA negative, they get benefit from the immunotherapy because not everybody is cured. Of course, the ctDNA positive, they get more benefit. This is very interesting because, finally, we have a flat curve here. Look at this. It’s flat. But what we did here is you mixed it—the patients that have ctDNA clear, this ctDNA clear—plus the complete pathological response. And now, finally, we can tell the patient, hey, if you have a complete pathological response and your ctDNA clears, you're cured. And maybe you don't need treatment. But, in this case, we're using two biomarkers. You know, two biomarkers because one biomarker is not enough. So, that's why maybe that's the solution for now. Until we get out the perfect technology, we can use two biomarkers. And now, finally, we have what we like, you know? If you are ctDNA negative, you are cured. And then, you don't need adjuvant therapy. And the same in this case. This is the CHECKMATE 816. Remember, three cycles of chemoimmuno with nivolumab before surgery. Same thing, you know, either ctDNA clears and you get chemo, you had the best group, but still it’s not 100%. It’s not 100%. And this is very interesting because this is what Julia Rotow presented at ASCO this year. Another way to use MRD is to escalate or de-escalate therapy. So, if these patients are patients that are in stage IV lung cancer, pretty much everybody gets chemoimmuno, you know? Nowadays, if the PD-1 is more than 50%, some people will give only immuno alone, but most of us give chemoimmuno. So, what she did here and she presented is that, basically, we are going to escalate the patients that don't—she gives pembro alone, and then she escalates the people that—so, the people that have a good response radiologically and a good plasma response, they stay on pembro alone. The people that have not a good response in radiology and people that had a—didn’t clear, they go for chemo. So, doing this—and, of course, the people that have the best response, have the best outcomes, you're skipping and saving chemo to a group of patients. So, instead of giving chemoimmuno to everybody, with this strategy, finding which patients have a plasma response and which patients have a radiologic response, we're sparing chemo in a group of patients. So, that's a positive outcome, It's a positive way to show that there is a value for MRD, okay? The other outcome is, oh, let's improve survival with MRD. Okay, that's a tough one, as you show. But at least we can spare chemo to the patient and get the same outcomes. That can be another way that we can use MRD until we can improve survival, no? And this is a design of how the perfect study will be. When you're MRD positive, you get therapy. When you're MRD negative, you don't get therapy. Thank you.