HR-Positive Breast Cancer: Predictive Biomarkers

Transcript:

Joyce A. O’Shaughnessy, MD: What did you guys think? That was the first time I had seen something that said there is a subset that doesn’t benefit from the CDK4/6 inhibitor in PALOMA-3, with gene expression profiling on the primary tumors. This is metastatic disease. This is primary tumor, cyclin E overexpression. That group really didn’t get much benefit from palbociclib. This is retrospective hypothesis generating, right?

Komal Jhaveri, MD, FACP: That’s what I think Nick Turner’s data showed in the metastatic setting, right?

Joyce A. O’Shaughnessy, MD: Yes.

Komal Jhaveri, MD, FACP: When they looked at the metastatic population from PALOMA-3 and the tumor samples, those who had a low cyclin E1 expression had a higher progression-free survival benefit compared with those who had a higher baseline cyclin E1 expression—here, we saw a lower benefit. That was interesting.

Joyce A. O’Shaughnessy, MD: It makes sense.

Komal Jhaveri, MD, FACP: Yes, it makes sense. It matches the biology.

Joyce A. O’Shaughnessy, MD: Given the mechanism of action of the CDK4/6 inhibitors and that cyclin E1 is more of CDK2, it makes sense that if you’re signaling through an alternative pathway, you may not benefit. So, this is interesting. It is intriguing, and it might help us in the long run, but it is still hypothesis generating, obviously.

Hope S. Rugo, MD: I think it would be interesting if you could look at levels of gene expression and be able to assess the benefit in the long run. Maybe we’ll be able to in collected trial populations or in the adjuvant setting. I don’t think we know if cyclin E1 changes with progression of disease. Presumably, it does, but I don’t think we have a lot of data on that. I think it would be fascinating to look at it in the adjuvant trials.

Debu Tripathy, MD: We have known that cyclin E isoforms that are cytoplasmic predict a worse outcome. If you look to see what escape mechanisms tumors can use, RB loss is one of them. We know that intact RB generally is required for CDK4/6 inhibition, but there are other cyclins that affect other parts of the cell cycle—E being the most prominent. So, eventually, these will turn up. The cyclin E isoform isn’t specific to CDKs but, as time goes on, I’m sure we’ll identify more and more. Fortunately, all of the trials did have tissue collection protocols, and we’re seeing the first generation of these come out now. There’s going to be much more as next-generation sequencing is done and as more deep sequencing is done on the blood samples. We look forward to more information.

Hope S. Rugo, MD: We’ll then know if cyclin E1 is altered with treatment. The other thing is that it’s associated with CDK2, right? There are now phase I trials with inhibitors of CDK2, 4, and 6, with the idea that these would overcome some of the resistance. But I think the Nick Turner data are really the only data we have that suggest there might be something that would predict for response.

Joyce A. O’Shaughnessy, MD: Yes.

Komal Jhaveri, MD, FACP: I think the other thing to probably see, which was interesting, as well, were the data that he presented at this ASCO Annual Meeting for PIK3CA and ESR1 mutations from the same data set. The ctDNA analyses, the paired samples that they looked at—it was interesting that patients who had a prolonged response and then ultimately progressed had emerging acquired PIK3CA, ESR1 Y537S mutations.

Hope S. Rugo, MD: It wasn’t a CDK4/6-specific effect.

Komal Jhaveri, MD, FACP: It could perhaps be related to fulvestrant, and there are preclinical data suggesting that Y537S might not be responding to fulvestrant as much when compared with the D538G mutation or the other...

Debu Tripathy, MD: Each 1 of these mutations is different.

Komal Jhaveri, MD, FACP: Exactly.

Hope S. Rugo, MD: The other interesting data looking at progression in early, first-line studies suggested that you get that ESR1 mutation, but then some of the patients lose their mutation over time. PIK3CA mutations were preserved, which is interesting. This would suggest that there’s a role for sequential therapy, even going back to exemestane/everolimus after second-line CDK4/6 inhibition.

Joyce A. O’Shaughnessy, MD: That’s interesting.

Hope S. Rugo, MD: Some of the patients will lose their ESR1 mutation.

Joyce A. O’Shaughnessy, MD: Yes.

Komal Jhaveri, MD, FACP: If it’s by ctDNA, we might have to be careful if they are subclonal. But PIK3CA are more clonal mutations. Perhaps that’s why they persist. So, it depends. We probably have to keep a close eye on whether it’s ctDNA or tumor, but they are definitely very intriguing. PALOMA-3 has generated so much of this data, which has been very interesting. What was also interesting that caught my eye was Nick Turner’s presentation about the decline in the PIK3CA-mutant allele fraction—the ctDNA ratio between day 1 and day 15 after you initiated a CDK4/6 inhibitor. And if you have the decline, you might respond to therapy if you are PIK3CA mutant. But, if you do not have that decline in the PIK3CA-mutant allele fraction by day 15, perhaps you’re not going to respond with palbociclib. I thought that was interesting.

Debu Tripathy, MD: It was a pretty good correlation with...

Komal Jhaveri, MD, FACP: To develop adaptive designs around that.

Hope S. Rugo, MD: It’s a subset and, as you said, cell-free DNA. That one, I think, is probably the weakest correlation.

Joyce A. O’Shaughnessy, MD: Yes.

Komal Jhaveri, MD, FACP: I agree.

Joyce A. O’Shaughnessy, MD: But it’s really good that the CDK4/6 inhibitors were not bringing out differential ESR1 and PIK3CA mutations. That was reassuring to me when I saw that.

Hope S. Rugo, MD: And the RB is low.

Joyce A. O’Shaughnessy, MD: Oh, very low. It was 4.8%, or something like that.

Transcript Edited for Clarity