Select Topic:
Browse by Series:

Biomarkers for CDK4/6 Inhibitors in HR+ mBC

Panelists: Adam M. Brufsky, MD, PhD, University of Pittsburgh Cancer Institute; Michael Gnant, MD, Medical University of Vienna; Joyce A. OShaughnessy, MD, Baylor University Medical Center; Hope S. Rugo, MD, University of California San Francisco Helen Diller Family Comprehensive Cancer Center; Michael Untch, MD, Helios Klinikum Berlin-Buch
Published Online: Friday, Oct 06, 2017



Transcript:

Adam M. Brufsky, MD, PhD: Let’s start with Hope, why do you think we can’t find a biomarker?

Hope S. Rugo, MD: I think we’re not smart enough. I actually think that we need better computational help with understanding biomarkers, that it’s not a single biomarker and it’s like the coagulation pathway, coagulation cascade. We have to think about everything as being a single pathway that’s activated. I think we understand now that there’s huge heterogeneity, there’s overlap. Just talking about, for example, if you have a PI3-kinase mutation, is it an activating mutation? Is it something different? How do you know whether it’s the right one or the wrong one? We’ve identified some, but there’s a lot more heterogeneity here. I actually think that looking at a combination of genes that are altered is going to be much more effective and really is the path to the future. And it may be that we need to be looking at a combination of genes and proteins. It just turns out to be really complicated. You need a lot of tumor tissue to do those analyses. It’s not like sending off a genomic sequencing for our patients, you actually need quite a bit more tumor. So, I’m actually very hopefully that being able to look potentially at a combination of blood and tumor will help us in that regard, and entering more into this idea of throwing a bunch of genes into a program that can try and create interrelationships that we can’t see ourselves.

Adam M. Brufsky, MD, PhD: So, you think it’s just more of a computational issue as opposed to there may be no biomarker?

Hope S. Rugo, MD: Well, I think we’ve established that and we’re not going to see any difference with that, there’s not a single biomarker. There’s not a clinical biomarker and there’s not a single genomic biomarker. But a combination looking at some kind of signature, I think maybe it’s gene and protein, gene, etc. You’re frowning, Joyce.

Joyce A. O’Shaughnessy, MD: Well, I’ll tell you the way I think of this CDK4/6 inhibition in the luminal A’s. So, the luminal B’s, they have some intrinsic resistance to endocrine therapy. One could see where CDKs could really help. But in the luminal A’s, of course cyclin D is an important transcriptional product of the estrogen receptor. I look at endocrine therapy plus the CDK4/6 inhibition as basically a double clamp down on the same extremely essential pathway. And our endocrine agents, even really good aromatase inhibitor inhibition or really good degradation of ER, they’re still not perfect in terms of really totally clamping down on that pathway.

So, to me, wherever ER is really driving, you’re going to get better clinical results from a dual inhibition on this critically important pathway. But the issue, what if you took a patient whose computational algorithm said was going to do really well with just good fulvestrant first-line? And let’s say that patient, like in the FALCON trial, gets the 22 months of this median progression-free survival because they’re nonvisceral. Then you came in second-line in those patients with the CDK4/6, would they ultimately be just as well off basically? To me, that’s more the question because I think the CDK4/6 is just another great way to get down that ER pathway in the luminal A breast cancer.

Hope S. Rugo, MD: I’ve 2 issues with that. I agree with you, but I think that tumors change over the course of progression in metastatic setting. So, saying that somebody has a luminal A tumor, yes, sometimes it’s conserved, but these tumors change over time. They become more proliferative, they respond to chemotherapy, but they don’t respond to hormone therapy. And if you look at intrinsic subtyping, although we don’t have tons of data on that, it does look like there’s a shift. Even from the beginning of neoadjuvant therapy to the end of neoadjuvant therapy, there’s a shift, and we believe that probably continues through time. So, that’s one thing.

Joyce A. O’Shaughnessy, MD: Because you’re worried about waiting on the CDK until second-line therapy because it may change.

Hope S. Rugo, MD: Right, that you’re changing. The other thing is that what we know, which is so interesting, is that people who have been on treatment for a long time, then they don’t respond as well until the next treatment unless we find some great targeted agent, trastuzumab or something else. But there’s some accumulation of resistance that occurs even to agents that you’re not seeing while you’re on treatment. And then the last thing is that it’s important for us to identify resistance mechanisms for the CDK4/6 inhibitors or sensitivity markers. Because if somebody gets 1 first-line with hormone therapy, they do actually respond to second-line hormone therapy. So, it doesn’t mean that you’ve developed resistance overall.

Michael Gnant, MD: I believe you touch on 2 important issues. First, we are victimized by our success actually, because we never saw this 10 years ago. We now have some of these patients with luminal disease. The median survival goes up to 50, 60, to 70 months, with individuals lasting for 15 years and longer. So, clearly the disease, as you have stated, changes and we see this. In the PALOMA trial, the ESR1 upregulation, these are things that the enemy can change its face somehow, also in a genetic level, which is posing a whole lot of new diagnostic challenges: how to detect this early, how you put patient on an investigational targeted drug early before there is massive explosion of clinical disease.

On the other hand, I would like to challenge you a little bit on that cumulation of resistance. Because, yes, we see these in-patient patterns. Systematically in the trials so far, we still haven’t really seen that. When you look, for example, at the everolimus, the BOLERO-2 trial, that was 4 and a half months’ PFS, which eventually lasted in total survival, not being significant any longer for statistical reasons. We have data from the PALOMA-2 trial that the time to the end of subsequent lines of treatment is even better in those patients who are on combination therapy as compared to the AI alone. While we are concerned that we would basically lose efficacy of subsequent treatments, including chemotherapy in endocrine responsive disease, I believe so far, we have been able to add on to the overall survival eventually, which I think is very reassuring.

Hope S. Rugo, MD: I think maybe as we control disease better we get less of that. And also, the people who respond the best have less genomic instability, so they get less of that, too.

Transcript Edited for Clarity 

Slider Left
Slider Right


Transcript:

Adam M. Brufsky, MD, PhD: Let’s start with Hope, why do you think we can’t find a biomarker?

Hope S. Rugo, MD: I think we’re not smart enough. I actually think that we need better computational help with understanding biomarkers, that it’s not a single biomarker and it’s like the coagulation pathway, coagulation cascade. We have to think about everything as being a single pathway that’s activated. I think we understand now that there’s huge heterogeneity, there’s overlap. Just talking about, for example, if you have a PI3-kinase mutation, is it an activating mutation? Is it something different? How do you know whether it’s the right one or the wrong one? We’ve identified some, but there’s a lot more heterogeneity here. I actually think that looking at a combination of genes that are altered is going to be much more effective and really is the path to the future. And it may be that we need to be looking at a combination of genes and proteins. It just turns out to be really complicated. You need a lot of tumor tissue to do those analyses. It’s not like sending off a genomic sequencing for our patients, you actually need quite a bit more tumor. So, I’m actually very hopefully that being able to look potentially at a combination of blood and tumor will help us in that regard, and entering more into this idea of throwing a bunch of genes into a program that can try and create interrelationships that we can’t see ourselves.

Adam M. Brufsky, MD, PhD: So, you think it’s just more of a computational issue as opposed to there may be no biomarker?

Hope S. Rugo, MD: Well, I think we’ve established that and we’re not going to see any difference with that, there’s not a single biomarker. There’s not a clinical biomarker and there’s not a single genomic biomarker. But a combination looking at some kind of signature, I think maybe it’s gene and protein, gene, etc. You’re frowning, Joyce.

Joyce A. O’Shaughnessy, MD: Well, I’ll tell you the way I think of this CDK4/6 inhibition in the luminal A’s. So, the luminal B’s, they have some intrinsic resistance to endocrine therapy. One could see where CDKs could really help. But in the luminal A’s, of course cyclin D is an important transcriptional product of the estrogen receptor. I look at endocrine therapy plus the CDK4/6 inhibition as basically a double clamp down on the same extremely essential pathway. And our endocrine agents, even really good aromatase inhibitor inhibition or really good degradation of ER, they’re still not perfect in terms of really totally clamping down on that pathway.

So, to me, wherever ER is really driving, you’re going to get better clinical results from a dual inhibition on this critically important pathway. But the issue, what if you took a patient whose computational algorithm said was going to do really well with just good fulvestrant first-line? And let’s say that patient, like in the FALCON trial, gets the 22 months of this median progression-free survival because they’re nonvisceral. Then you came in second-line in those patients with the CDK4/6, would they ultimately be just as well off basically? To me, that’s more the question because I think the CDK4/6 is just another great way to get down that ER pathway in the luminal A breast cancer.

Hope S. Rugo, MD: I’ve 2 issues with that. I agree with you, but I think that tumors change over the course of progression in metastatic setting. So, saying that somebody has a luminal A tumor, yes, sometimes it’s conserved, but these tumors change over time. They become more proliferative, they respond to chemotherapy, but they don’t respond to hormone therapy. And if you look at intrinsic subtyping, although we don’t have tons of data on that, it does look like there’s a shift. Even from the beginning of neoadjuvant therapy to the end of neoadjuvant therapy, there’s a shift, and we believe that probably continues through time. So, that’s one thing.

Joyce A. O’Shaughnessy, MD: Because you’re worried about waiting on the CDK until second-line therapy because it may change.

Hope S. Rugo, MD: Right, that you’re changing. The other thing is that what we know, which is so interesting, is that people who have been on treatment for a long time, then they don’t respond as well until the next treatment unless we find some great targeted agent, trastuzumab or something else. But there’s some accumulation of resistance that occurs even to agents that you’re not seeing while you’re on treatment. And then the last thing is that it’s important for us to identify resistance mechanisms for the CDK4/6 inhibitors or sensitivity markers. Because if somebody gets 1 first-line with hormone therapy, they do actually respond to second-line hormone therapy. So, it doesn’t mean that you’ve developed resistance overall.

Michael Gnant, MD: I believe you touch on 2 important issues. First, we are victimized by our success actually, because we never saw this 10 years ago. We now have some of these patients with luminal disease. The median survival goes up to 50, 60, to 70 months, with individuals lasting for 15 years and longer. So, clearly the disease, as you have stated, changes and we see this. In the PALOMA trial, the ESR1 upregulation, these are things that the enemy can change its face somehow, also in a genetic level, which is posing a whole lot of new diagnostic challenges: how to detect this early, how you put patient on an investigational targeted drug early before there is massive explosion of clinical disease.

On the other hand, I would like to challenge you a little bit on that cumulation of resistance. Because, yes, we see these in-patient patterns. Systematically in the trials so far, we still haven’t really seen that. When you look, for example, at the everolimus, the BOLERO-2 trial, that was 4 and a half months’ PFS, which eventually lasted in total survival, not being significant any longer for statistical reasons. We have data from the PALOMA-2 trial that the time to the end of subsequent lines of treatment is even better in those patients who are on combination therapy as compared to the AI alone. While we are concerned that we would basically lose efficacy of subsequent treatments, including chemotherapy in endocrine responsive disease, I believe so far, we have been able to add on to the overall survival eventually, which I think is very reassuring.

Hope S. Rugo, MD: I think maybe as we control disease better we get less of that. And also, the people who respond the best have less genomic instability, so they get less of that, too.

Transcript Edited for Clarity 
View Conference Coverage
Online CME Activities
TitleExpiration DateCME Credits
Publication Bottom Border
Border Publication