Adam Brufsky, MD, PhD: That’s the OlympiA [trial], and I think the next big thing is the monarchE [trial], which I believe was presented at San Antonio [Breast Cancer Symposium] 2020, and Nadia Harbeck [MD, PhD,] has published, was it in the New England Journal of Medicine?

Komal Jhaveri, MD, FACP: It was in the Annals of Oncology, given that it was presented at ESMO [European Society for Medical Oncology annual meeting].

Adam Brufsky, MD, PhD: It was presented at ESMO. The thing is that in this trial, it was patients with hormone receptor-positive and HER2-negative, node positive, high-risk early stage breast cancer, and the way high risk was defined I think is the crux of this particular study. But what happened was, you had either 4 nodes positive, or 1 to 3 nodes positive with at least a histologic grade 3 or a T3 tumor. For the T3 N1, it makes you stage III. But histologic grade 3 with a T2 tumor, stage II tumor, is a really interesting one that we can talk about. There was another cohort of that, that inclusion based on Ki-67. And you could have, again, N1 disease, but have a Ki-67 greater than 20% without a high grade and without T3 disease. An absolutely huge number of patients, 5600, were randomized 1:1 with the standard of care endocrine therapy with or without abemaciclib for 2 years.

It was a really interesting and very straightforward trial. In the Ki-67–high subgroups, with a median of a little more than 2 years of follow-up, there was a separation of the curves, and basically, there was an absolute improvement in 3-year invasive disease-free survival rates of about 6% in the entire intent-to-treat cohort. In the Ki-67–high population in cohort 1, that’s the cohort that was based on clinical parameters, you can see there was something similar, a 7.1% 3-year invasive disease-free survival. Thus, clearly, this is something that has a benefit, a substantial benefit, in high-risk patients. I think our problems and where there has been a lot of discussion, as you know, over the last year, year and a half, after this was presented, there are 2 issues. The first one has to do with how do you define high risk? My own personal bias is basically to do stage III. I’m not somebody who will do all the stage IIs, N1 disease, with a high Ki-67. I don’t know how you feel about that. Will you give it to a stage II patient with a high Ki-67?

Komal Jhaveri, MD, FACP: No, I think there are 2 important discussions we need to have. The first one is the Ki-67 itself, and how the approval came about in that patient population, when that was cohort 2. Cohort 1 was one without the Ki-67 information. That’s one, and we should talk about that for sure. The second is how comfortable do we feel that about a T2 tumor with N1 to N3 disease versus the highest risk patient population, which is 4 or more nodes, given the absolute benefit that you described as about 5%, 6%? And we do know there is some toxicity involved. I have to say, it depends. I have definitely done that even in a T2 tumor. I think the younger patients are very motivated and very keen, and now even very knowledgeable about this information and this approval. Thus, I have been doing that, even for those patients. But let’s talk about the Ki-67. Do you do that routinely at your institution?

Adam Brufsky, MD, PhD: Yes. In our institution, we have really high-quality and high-powered immunohistochemists. A guy, David Dabbs, [MD,] one of our old pathologists who retired, wrote the book on immunohistochemistry, actually, one of the classic texts. Thus, we’ve been doing Ki-67s for 20 years, and we use it to stratify patients up front for neoadjuvant therapy, believe it or not, especially the ER [estrogen receptor]-positive patients. For someone who comes in with a Ki-67 above 20% or 30%, we generally will use neoadjuvant therapy in the ER/PR [progesterone receptor]-positive patients. We’re very familiar with it. The Magee Equations that we use have Ki-67 as one of the parameters, so we’re very familiar with it.

But I think what you’re getting at, if you don’t have a pathologist who’s comfortable with Ki-67, they have to be trained again on it to be more comfortable with it. I think that’s the issue with this. I think that’s one of the reasons we were all a little surprised that the FDA gave the approval based on high Ki-67. It’s a test that a lot of people can do; I’m sure your pathologists and most academic pathologists are really good, and very quickly will become proficient at it. But I think going outside of the academic pathologist, maybe someone who does 20, 30 cases a year, maybe 40, is somebody who may not necessarily be comfortable with Ki-67. How do you handle that? You guys were not doing it routinely, is my understanding at Memorial [Sloan Kettering Cancer Center], is that correct?

Komal Jhaveri, MD, FACP: That’s absolutely correct. We were not doing that routinely, and we can request this now based on these data that have come out, for a given patient. But yes, at the point you were coming at also was that the FDA label was for patients who were part of the cohort 1 that you pointed out, the 4-plus nodes, or 1 to 3 nodes with either a larger tumor or a higher grade, but they also had a Ki-67 of 20% or higher added to that for the approval. Then we also saw the ASCO [American Society of Clinical Oncology] guidelines and the EMA [European Medicines Agency] guidelines that came out shortly after that followed the trial patient population regardless of the Ki-67. The paper the FDA published in the JCO [Journal of Clinical Oncology], trying to justify why they did what they did, was very interesting as well because they point out that when they looked at the overall survival data, the hazard ratio there for the intent-to-treat cohort 1 patient population, or the intent-to-treat population, was closer to 1.

But I think the benefit was slightly larger in the Ki-67–ؘhigh patient population, which is why they thought that given the benefit, given this guarded follow-up of 28 months, and we think about ER-positive disease as a late recurrence. And is this addition of therapy truly just delaying recurrence, or is it eliminating recurrence to address that? I think that’s why they came up with this label. I have to tell you though, I have been following the ASCO and the EMA guidelines and the monarchE trial criteria, and not necessarily depending on the Ki-67, unless I’m really trying to make a case for a given young patient who might have say 2 nodes and a 3.5-cm tumor, but grade 2. That would have not fit the bill, and I would try and get the Ki-67 to make that determination for that given patient. But otherwise, I have been following the ASCO and the EMA guidelines.

Adam Brufsky, MD, PhD: Yes, me as well. I’m following those guidelines as well. But again, I’m not as convinced that a stage II with a high Ki-67 has that high of a response. You’re always trying to figure out the absolute benefit to this, and really the stage IIIs are the patients I tend to use it on. Before we move on from this, it surprised me that the PALLAS and PENELOPE-B trials were negative. Again, with a CDK4/6 inhibitor that has the same PFS [progression-free survival] as abemaciclib in the metastatic setting, both in the first line and second line. It’s interesting that we didn’t see that in the adjuvant setting. Do you have any ideas of why that is?

Komal Jhaveri, MD, FACP: I think there are so many reasons that we can try to tease and look into. First of all, I think the patient populations were slightly distinct between the monarchE and say the PALLAS study. I think the PALLAS study did have a decent proportion of the stage II patient population that you were just pointing out, that might not necessarily have that good absolute benefit. In monarchE, 60% were N2. This was definitely the highest risk patient population in this trial compared to that one. When we think about the drug itself, in the metastatic setting, we haven’t seen a big clinical difference when it comes to progression-free survival. Having said that, abemaciclib is a continuous dosing drug. It also has approval as monotherapy based on the MONARCH 1 data in the metastatic setting. Could that be one of the potential reasons has always been a speculative hypothesis for that. Then, the third thing was did patients really take the drug enough, and did that make a difference in terms of the discontinuation rate? But there was a recent publication that came out from the PALLAS group that they did not really see a big difference in the patients who discontinued versus who did not discontinue the drug. I think the jury is out as to why exactly we saw what we saw, but here we are.

Adam Brufsky, MD, PhD: Yes, I agree with you.

Transcript edited for clarity.