SERENA-6 at ASCO 2025: ctDNA-Guided Therapy in HR+/HER2- mBC

In this episode, Hope Rugo, MD, of City of Hope, shares insights from the SERENA-6 trial, which evaluated ctDNA-guided treatment changes for patients with HR+/HER2-negative metastatic breast cancer. She explains the reasoning behind monitoring for ESR1 mutations before clinical progression and details how switching to camizestrant with a CDK4/6 inhibitor impacted progression-free survival compared with continuing an aromatase inhibitor. According to Dr. Rugo, these results could mark a turning point toward earlier, biomarker-driven treatment adjustments for patients with HR+/HER2-negative breast cancer.

Transcript

Dr. Rugo: Hello, I’m Hope Rugo, a breast medical oncologist and director of the Division of Breast Oncology and the Women’s Cancers Programs at the City of Hope Comprehensive Cancer Center. At ASCO 2025 this year, we heard the much anticipated results of SERENA-6. I’m going to start with a few different areas to cover explaining SERENA-6, what brought the trial into being, what the drugs are that are being used and what the implications are for this trial and for other trials like this in the future. So the first question is really why was SERENA-6 trial initiated as a ctDNA-guided trial? The whole idea behind the trial is based on the idea that you could find molecular evidence of disease becoming resistant to current therapy and that if you targeted the cancer before there was imaging or symptom evidence of progressive disease that you would have a better response. There would be less burden of disease. There would be less resistance that has developed under the pressure of treatment.

A previous study that was investigator initiated called the PADA-1 trial suggested that if you changed hormone therapy at the time of detecting an estrogen receptor mutation in the blood through circulating tumor DNA, or ctDNA, that you would improve disease control compared to staying on the same endocrine therapy. And indeed, we know that if you change therapy, you will have a longer time of disease control than if you keep on the same therapy, even if you didn’t know that you had an estrogen receptor mutation or an ESR1 mutation. So the results of SERENA-6, the whole trial design was really set up to try and investigate this in a larger population in a prospective manner with an understanding of how we’re benefiting patients and whether changing therapy based on this molecular evidence of developing resistance can make a difference.

So it’s a complicated question, and I think the idea is that we already know that if CT levels are going up, so the variant allele fraction, or VAF, is increasing while a patient is being treated versus somebody who you can’t find the ctDNA or it’s very low levels that their time to disease progression will be shorter, those that have the increasing ctDNA and blood. So we already know that it’s an indicator, and we had explored this in the past but not in this way. And I’ll explain that in just a moment. But looking at, for example, circulating tumor cells, if the circulating tumor cells were increasing or they hadn’t decreased with treatment, the question was, in a randomized trial, if you change treatment, would you get a better outcome? Would patients have lived longer with their cancer? Would they have better disease control? And the answer was no, and that was a little discouraging. And I think part of the reason for this at least we thought at the time, was that circulating tumor cells really detected escape from treatment or lack of disease control too late.

It was too gross a measure and also, some people just don’t have circulating tumor cells and their cancer is still growing. And then lastly, and maybe most importantly for the results we found with SERENA-6 is that we didn’t know what treatment to change them to that could overcome the resistance of the higher circulating tumor cells. Prior to looking at circulating tumor cells, we looked at tumor markers, and when the tumor markers go up, we learned not to change therapy. We wait for either symptoms or imaging evidence of disease progression in patients with metastatic disease because it didn’t help patients to change only based on rising tumor markers. And we were worried that we would run out of treatment too quickly. So what SERENA-6 did was to look at evaluating patients who have metastatic breast cancer hormone receptor positive HER2-negative and monitor them on a regular basis to see when their circulating tumor DNA showed evidence of an ESR1 mutation.

We know from randomized trials in the past that looked at ESR1 mutations that aromatase inhibitors don’t work well if you have an ESR1 mutation. We know from studies that have monitored patients and monitored the blood for ctDNA that when you first get diagnosed with metastatic breast cancer that the rate of ESR1 mutations is less than 5%. But for patients who are on an aromatase inhibitor over time that up to 40% of patients will have evidence of an ESR1 mutation by the time their cancer is progressing. So this is a mutation unlike many others that we follow that develops under the pressure of treatment, so we knew that that was the case. Then studies looked at patients who were treated with the aromatase inhibitor, exemestane, versus the injectable selective estrogen receptor down regulator, or SERD fulvestrant and found that although if you looked at the population as a whole, the exemestane, the aromatase inhibitor and fulvestrant looked like they worked about the same way.

If you looked at the patients who had an ESR1 mutation, you found that only fulvestrant was really working in any clinically meaningful way. So the aromatase inhibitors didn’t work. Sorry. So then the question was, does fulvestrant always work well? Well, we found that there are certain ESR1 mutations where fulvestrant doesn’t work as well. And we’ve now seen in two randomized Phase 3 trials that an oral SERD as opposed to the injectable SERD fulvestrant work better, control the disease better in patients with metastatic breast cancer who have evidence of ESR1 mutations. So that’s been seen with the approved oral SERD elacestrant and as still experimental oral SERD imlunestrant. So camizestrant is the third oral SERD to have Phase 3 data. Camizestrant was studied in a Phase 2 trial that showed that it also worked better than a fulvestrant in patients who had evidence of an ESR1 mutation, and the ESR1 mutation in blood following ctDNA seemed to clear rapidly when you used camizestrant versus fulvestrant.

So based on the PADA-1 trial and based on the idea that maybe if you could change therapy based on molecular evidence of resistant disease that you could improve outcome for patients, SERENA-6 was developed. So camizestrant would work better than certainly an aromatase inhibitor in patients who had molecular evidence of a estrogen receptor mutation by ctDNA. Now one of the things that SERENA-6 wanted to do that was a little different than that PADA-1 trial, which was a smaller trial, was two different things. So one was to try and limit the patients who have true endocrine-resistant disease and develop progression of disease within six months of starting their aromatase inhibitor and CDK4/6 inhibitor. So they started screening for ESR1 mutations after patients had already been on for six months. And when patients signed consent, they had to do CT scans and other imaging as appropriate to make sure that they didn’t already have evidence of progression of their cancer that was asymptomatic. The second thing that SERENA-6 did in order to try and make the best evaluation, the clearest evaluation of the effectiveness of this approach was to make the trial Phase 3 and double-blind.

So that meant when patients went on to camizestrant versus staying on their aromatase inhibitor, they would get another placebo pill so they wouldn’t know what treatment they were on. And that’s really important I think, in a trial like this. And camizestrant, of course, had the very nice Phase 2 data showing that it worked well in patients with ESR1 mutations. So the first thing is to screen patients. So they would screen patients for ctDNA, make sure they didn’t have evidence of metastatic disease, and then they would continue to draw blood every two to three months to look for an ESR1 mutation. At the same time, because we do scans to look for disease progression, patients would have scans on a regular basis every eight weeks to make sure they didn’t already have evidence of progression of their cancer. And if they did, of course, they were off trial because this study wanted to change treatment before a disease progression. So a total of 3,256 patients started the surveillance. And interestingly, this is 40% of patients will develop this over time, but there could be a long time until you get an ESR1 mutation.

We don’t know how long it will take. So at the time that surveillance closed and they’d already enrolled all the patients they needed, there were 1,949 patients that were still ongoing in surveillance; no evidence of disease progression and no evidence of an ESR1 mutation. So it’s important to keep that in mind. During the course of screening, they found 548 patients who had an ESR1 mutation. Half of those patients had a mutation on their first test, about a little under 40% after two to five tests. So there is a group of patients, we’re not sure who those are, who have ESR1 mutations that develop relatively early in the course. But remember, patients had to be at least six months after starting their aromatase inhibitor and CDK4/6 inhibitor, but they could have been two years or three years on therapy before they joined into the trial. And I’ll tell you more about that in just a moment. So the trial was designed to continue the same CDK4/6 inhibitor that the patient was on when they started screening, regardless of which arm they randomized on to.

So patients knew they would be continuing their same CDK4/6 inhibitor and just changing the endocrine therapy agent, but they wouldn’t know the treatment had changed because it’s double-blind, a placebo pill and the hormone pill but they look the same and you don’t know which pill is which arm you were randomized to. So out of the 548 patients who had a detection of an ESR1 mutation, 233 discontinued. Now, one of the issues for failing is that they would have been found to have disease progression, but that was only 53 patients. But other eligibility criteria played a role and then for a lot of patients we don’t really know, and there’s a lot of effort going on to try and figure out why they weren’t randomized. So 315 patients were randomized, which is what they were looking for in the trial itself. I think that it’s really amazing that patients can come in and get blood drawn every two to three months to look for evidence of this developing molecular resistance. But we still have the big question of whether or not that would be the right approach for patients.

So SERENA-6 first looked at how long it was until the disease started to progress again after… to progress, not again because they didn’t have progression, but to progress after they’re randomized. And that’s really the primary endpoint of the trial. There were factors that were put into the trial to allow the key factors to be balanced between the two arms. And one of the key factors was, how long ago did you start the aromatase inhibitor and CDK4/6 inhibitor until you were randomized? So they wanted to know early versus late, less than 18 versus 18 or more months. And then also, of course, we want to know which CDK4/6 inhibitor patients were on, so that was balanced in the arms, and you wanted to know if patients had visceral or non-visceral disease because visceral disease tends to correlate with less hormone sensitivity. And then lastly, I think one of the areas that we all think is important is how quickly does that ESR1 mutation develop?

As I said, almost 2000 patients who entered screening were still without an ESR1 mutation or disease progression at the end of study screening. So it’s going to be really interesting over time to know who those patients were versus the patients who had an ESR1 mutation quite early in the course of the disease. Now why choose camizestrant? Well, obviously this was in part the sponsor’s decision. Camizestrant is a very active oral SERD. We knew that from the Phase 2 trial, and we want an oral agent rather than injectable agent because we already know in patients with ESR1 mutations that these agents are superior. We also want to make sure that the agent is well tolerated and we already knew that camizestrant was a well-tolerated oral SERD with some unique side effects we’ll talk about in just a moment. So what did we learn? So this was a interim analysis and it just looks at the first primary endpoint, which is, how long did patients stay on either their continued AI and CDK4/6 inhibitor or camizestrant and CDK4/6 inhibitor in a blinded fashion until their disease started to grow?

So who were these patients? I think it was really interesting to see that the median time from start of AI CDK4/6 inhibitor to randomization was almost two years, 23 months. Because 50% of these patients had a positive test at first screen, that means that probably a lot of these patients were in that 1 1/2 year range. And in fact, 62% or so had started within 18 months. Sorry, I’ll record that again. 62% of these patients had started their AI and CDK4/6 inhibitor at least 1 1/2 years ago before they found the ESR1 mutation, so 18 months or longer in about 62 months… 62%. A little less than 40% were less than 18 months. So that makes a lot of sense why it was a 23-month median, as one would expect based on when the study was started, about 3/4 of the patients, 75% took palbociclib, 15% ribociclib, and 10% or less took abemaciclib. There also was pretty even distribution between the different ESR1 mutations, and that will be looked at in further analyses of this trial data. And about 80% of women were postmenopausal.

Now remember, these are patients who are on their first line therapy, so about 43% of patients had visceral metastasis, so really interesting patient population. So what happened? Well, the patients who switched to camizestrant had a markedly longer progression-free survival, so until their disease progressed than the patients who were on an AI. If you were on an AI a median of 23 months, the median time to progression-free survival from randomization was an additional nine months. So that’s long, actually. This was a very endocrine-sensitive group of patients, less than 50% with visceral metastasis. But the patients who changed to camizestrant had control for an additional seven months, so 16 months longer. If you think about it, they were staying on the same CDK4/6 inhibitor. We don’t know that the same CDK after… the same CDK actually helps a whole lot. Switching helps. We’ve seen that with the postMONARCH and ENABLAR-3 data with imlunestrant and amebaciclib, but staying on the same CDK4/6 inhibitor, we don’t know how much it helps.

And if you just look at giving the individual oral SERD, we would expect in this patient population with ESR1 mutations that that PFS might not be quite as long. 16 months plus the 23 months, that’s a very long time. And the hazard ratio was 0.44, so that’s a 56% relative improvement. One of the interesting things we saw was that the curves started to separate around the time of the first scan. So scans were done every eight weeks. Just about two months is where you’re seeing those curves separate and they continue to separate over time. And one of the most impressive parts of the data to me was looking at landmark analyses at 12 and 24 months. At 12 months, double the number of patients were free from progression. But at 24 months, which is really far out, only 5% of the patients who were on the AI were free from progression. But 30% of the patients on camizestrant were free from progression. And I think one of the things we learn outside of the medians is what happens to these outlier populations?

How are we benefiting our patients in the best way? They looked at subset analyses to try and figure out what might be different, but all of the little dots were to the left of one with some of the confidence intervals crossing one, but not a lot. There were so few patients who were on some of these agents, it’s a little bit hard to know what to do with the subgroup analyses of progression-free survival. Even the number of patients, for example, treating Asia was relatively small. So that was a small group, a little bit hard to tell what was going on in that group of patients. But the benefit was seen across age, across timing of ESR1 mutation and regardless of whether it was early or late, based on that cutoff of 1 1/2 years from when you started your AI and CDK4/6 inhibitor and regardless of whether you had visceral or non-visceral disease. So I think that’s really helpful.

They showed something that was interesting and I think there’s a lot of questions about it, and that’s time to deterioration in global health status quality of life using a tool that’s considered a standard tool worldwide, the EORTC QLQ-C30 test. What that showed was that there was an immediate fall-off of quality of life in all patients, about 10% at the first test, the first forum, the questionnaire that people were filling out. We don’t know why that was true. And then the curve separated right away with a continued fall-off in the AI CDK4/6 inhibitor arm. Now was that because people expected that their arthralgias would improve and they didn’t because they didn’t know what agent they were on? So we don’t really know. And I think what we discussed at ASCO when we were talking about this remarkable data was that we really need to see all the domains.

Why is the time to deterioration so much better with camizestrant than in AI? The mean time to deterioration of at least 10 points or more was 6.4 months for an AI, about three months before the median PFS and 23 months for camizestrant, so really a very long time. So that’s, I think, quite remarkable and we need to investigate it better because it’s longer than the median PFS. So we don’t really know why that was. So we’ll see that data, I think, in more detail probably a little bit later this year. One of the big secondary endpoints that we all want to see as treating physicians is did this translate? And when people change to their next therapy, when they were on the AI CDK4/6 inhibitor, did they have a longer PFS? And we don’t have enough mature data yet on that. And we also need to know what treatment patients went on with the second PFS because, of course, the camizestrant people are now in their third treatment.

The AI and CDK4/6 inhibitor people are on their second treatment, so this is all going to be data for the future. In terms of safety, camizestrant was very well tolerated. Almost no one discontinued in one patient due to an adverse event, and the treatment exposure was much longer as you would expect from camizestrant. One of the side effects that had been reported earlier with camizestrant really wasn’t seen in 10% or more of patients, and that’s bradycardia, but they did see photopsy of flashing lights. It was mostly very low grade and it happens when people go from light to dark. It didn’t bother patients much. They didn’t discontinue for it, so that wasn’t an issue. There was neutropenia because 3/4 of patients were on palbociclib, but nothing that you wouldn’t expect already. And there really weren’t remarkable differences in any of the other factors that were seen.

Diarrhea occurs, but this was seen across both arms, so it wasn’t really any different and there wasn’t much nausea either. So they looked at the impact of photopsy of these flashes of light that go away rapidly on the ability to conduct daily activities, and there really no difference. So we end up with this really intriguing data where we see for the very first time in breast cancer that monitoring for evidence of molecular resistance and then changing therapy based on molecular resistance or not actually changes the progression-free survival from the time of randomization. So is this something we’re all going to adopt right away? I think that it’s intriguing. I would consider checking an ESR1 mutation in a patient where I think that their treatment may be starting not to work quite as well.

So median time about 23 months, so maybe a year and a half in a patient who I think is having a very nice response. I’m a little worried about checking for ESR1 mutations every two to three months, and we certainly don’t do scans every eight weeks in clinical practice. So I think that it will be interesting to see what we find in the real-world setting. Would you get the same results if you changed after disease progression? I think the concern is that you’ve developed more resistance to endocrine therapy, that you might have more symptoms that are hard to control. If you have progression in bone, does this mean that patients will have more bone pain over time because you waited to change therapy? But you don’t want to change very early either.

So more analyses of the data from SERENA-6 will be critical, I think, to understand how to apply this data in the best way in the real-world setting. Of course, we don’t have camizestrant approved and we don’t have this type of monitoring approved, so we don’t have to worry about putting it into routine clinical practice tomorrow. But it might be reasonable if you’re worrying that a patient is progressing. The scan, we have a lot of problems with patients with ER-positive disease where the scan doesn’t really tell us what’s going on. You see these tumor markers rising and you’re doing scans, or they have some symptoms and the scans don’t show anything because the disease doesn’t light up or it doesn’t really show up on CTs or bone scans. There, I think, ctDNA could be really helpful in understanding what’s going on. And if you do see an ESR1 mutation in a setting like that, changing therapy could be quite beneficial.

I hope that in the future we’ll have data doing this kind of approach. But changing the CDK4/6 inhibitor, I think that would be quite intriguing and understanding because I think now we like to use these drugs in combination. We only have one oral SERD approved today, that’s elacestrant, and it’s approved as a single agent. There is safety data in combinations from an ongoing ELEVANT trial. So you could certainly try a combination approach until we see approval of imlunestrant and camizestrant in combination, which I expect we’ll see in the next six months. So this is really exciting data. It’s a step forward for our patients and for all of us in understanding how to better treat patients with the most common subset of breast cancer, which is the most common cancer in women worldwide. The whole idea of being able to more effectively monitor the development resistance with ctDNA is intriguing, and I think that we’re going to be adopting some approach like this more in the future. Thanks so much for listening, and I hope this was helpful.