Advanced Breast Cancer: Key Updates in Clinical Care - Episode 4

Selecting a CDK4/6 Inhibitor in HR+ Breast Cancer

Transcript: Joyce O’Shaughnessy, MD: So what about choice of CDK [cyclin-dependent kinase], just as our practice patterns have evolved a bit? I think we all probably utilize all of the CKD4/6 inhibitors in our practice. But are there any comments to make? We did take a look at the MONARCH 2 and 3 subset analyses, in some detail, looking for the higher-risk patients versus the lower-risk patients. Now, of course, all of the trials, both in the first- and the second-line, if you look at the populations, the results are just absolutely superimposable.

But it’s interesting. We know the patients with the really virulent aggressive disease is a smaller percentage in the metastatic setting. So those patients won’t necessarily be reflected in the overall trial population. That’s the first thing to say. There are no comparison data. There is no head-to-head comparison. And all of the studies have done a subset analysis, and every single subset benefits across all disease characteristics. So we know this. But in the MONARCH 2 and MONARCH 3 studies looking at patients with liver metastases, grade 3 disease, who were progesterone receptor-negative, and also with short treatment-free intervals—so recurring within a short period of time of adjuvant endocrine therapy, or on adjuvant endocrine therapy—seeing robust response differences of 30% or more as well as improvements in PFS [progression-free survival].

Also, benefit in patients with more indolent disease: Bone only, progesterone receptor-positive, lower grade. They also benefit, but interestingly to a lesser extent. So I have sort of developed, in my mind, that if I have true luminal B—you know, a real virulent full throttle luminal B—I will utilize abemaciclib. Of course, I was involved in the subset analyses, so I’m very familiar with the data. But also, I think the continuous inhibition of CDK4/6: There may be a little bit of a broader spectrum against CDKs. Again, we don’t have a head-to-head comparison. We just have to make these clinical decisions based on our gestalt, if you will. I must say, too, it is also a bit influenced by the later-line patients who we treated in the beginning when we first had these agents. I had patients getting responses and having had everything, and they were on the abemaciclib and, significant liver disease. So I had some anecdotal experience too that did kind of sway me, in that regard.

And when it comes to more indolent luminal-A disease, I want to go for the least toxicity that I can get. And there I’m utilizing the palbociclib and ribociclib pretty much exclusively. So that’s what I have come down to with my practice. But how do you think about it, Ruth?

Ruth O'Regan, MD: I think you make some very valid points. I have tended to use palbociclib in the main, but your points certainly make me think about the patients who you have described. If patients have brain metastases, I do tend to go to abemaciclib because the data are stronger there. But other than that, I think they’re all very effective. We don’t have the same data you’re talking about with the other ones, so I don’t know.

Joyce O’Shaughnessy, MD: Yes.

Ruth O'Regan, MD: I do like your algorithm. I think that makes sense. But certainly for CNS [central nervous system], I would certainly think about abemaciclib. I think for a leptomeningeal disease, unfortunately we don’t know. I certainly would recommend abemaciclib for a patient with leptomeningeal disease, but I don’t know how well it would work, honestly.

Joyce O’Shaughnessy, MD: How about you, Kevin?

Kevin Kalinsky, MD, MS: I agree with Ruth’s comments. I also think what you were mentioning, in terms of that treatment-free interval, if you think about the MONARCH 3 subsets, which had that treatment-free interval—like a 3-year cutoff—those who were less than 3 years, which are really bad actors—patients who are on hormonal therapy and then recur within that time period—the difference in the Kaplan-Meier curves was quite notable with a hazard ratio that was less than 0.5. So I think your point is well made, and I can see where you’re coming from based upon those data.

Hope S. Rugo, MD: I think we have to make decisions, and I think that this is a nice way to parse it out. But I don’t know that the subset analysis really tells that there’s a difference between the agents. We did a subset analysis in the PALOMA-2 trial and looked at all of these different disease-free intervals, and there was no difference. I mean the hazard ratios all line up.

And one of the things about the MONARCH data is that abemaciclib works really well probably across all of the subsets because the medians weren’t reached in some of those really good prognosis subsets. So I wonder what we’re going to see in the longer follow-up, where there’s going to be still a really big difference favoring abemaciclib, even in the good prognosis group.

So when we look at these forest plots, basically the agents work across multiple different subgroups, prognostic subgroups, and I think that just like when we use the AIs [aromatase inhibitors], we have to decide how we’re going to give the drugs. So somebody who can’t really keep track of their pills—that’s somebody on capecitabine now—giving them continuous dosing with abemaciclib is good. For somebody who has a lot of neutropenia or bone marrow dominant disease, where they have low counts, abemaciclib is a really good choice. And maybe what you’re parsing out would be more visceral disease for abemaciclib or palbociclib for somebody who has bone-only disease. You have to make some decisions there. But I’m not totally convinced that there’s differential benefit.

Joyce O’Shaughnessy, MD: Yes, because their subsets haven’t been done exactly the same across all of the different studies. That’s the thing. So we can only look at the data we have.

Kevin Kalinsky, MD, MS: I also think that Hope’s comment, that subset that we’re talking about is a small number, right? There are less than 100 patients. So to really read significantly into it should be done with some caution.

Joyce O’Shaughnessy, MD: Yes. It’s interesting just to see a little snapshot of where we are in practice.

Transcript Edited for Clarity