Early-Stage HR+ Breast Cancer: Ki67 and the monarchE Trial

Transcript:
Aditya Bardia, MD, MPH: Continuing on this topic, [let’s have a] final discussion related to use of abemaciclib in the adjuvant setting. The FDA approval required 3 specific criteria: patients who had lymph node–positive disease; high-risk features, either more than 4 positive nodes or 1 to 3 positive nodes of grade 3; and Ki-67 of more than 20%. All 3 criteria needed to be met, from an FDA-approval perspective. But then ASCO [American Society of Clinical Oncology Clinical Practice Guidelines] and NCCN [National Comprehensive Cancer Network] Guidelines said that abemaciclib can be used based on the monarchE intent-to-treat analysis, which doesn’t require Ki-67 to be more than 20% if the other criteria are met. In your practice, how are you using abemaciclib? I’ll start with Heather.

Heather McArthur, MD: As a reminder, the monarchE study randomized 5637 patients with hormone receptor–positive, HER2 [human epidermal growth factor receptor 2]–negative, node-positive, high-risk, early stage breast cancer. It’s worth the reminder, especially in light of this discussion, because some data were presented that touch on the Ki-67 issue that we’ve been circling around. There were 2 cohorts in this study. The first cohort qualified based on high-risk clinical and pathological features—4 or more lymph nodes or 1 to 3 lymph nodes, and either grade 3 disease or a tumor greater than or equal 5 cm. There was a group that qualified that was Ki-67 agnostic. The second cohort incorporated high risk with the Ki-67—1 to 3 lymph nodes, Ki-67 greater than or equal to 20%, and grade and tumor size criteria. It’s worth refreshing our memories that there were 2 different cohorts.

With 42 months of median follow-up at San Antonio Breast Cancer Symposium [SABCS], investigators presented invasive disease-free survival benefit that persisted beyond completion. That’s almost 4 years of follow-up with a hazard ratio of 0.664, and distant relapse-free survival has a ratio of 0.659. Those curves didn’t collapse after completing abemaciclib therapy. In fact, the benefits seemed to deepen, and landmark analyses were presented at each of the years on study. In year 3, the hazard ratios for those 2 end points were more impressive than they had been in earlier years. That effect seems to deepen.

But what’s particularly germane to this conversation is the data that were presented around Ki-67. They showed Kaplan-Meier curves from cohort 1—that’s the cohort that qualified based on high-risk clinical and pathological features, so it was Ki-67 agnostic. They took patients in that cohort, where Ki-67 information was available, and presented Kaplan-Meier [curves] based on low vs high Ki-67. Having a low Ki-67 was favorable in terms of prognosis. But the magnitude of benefit was very similar for the low Ki-67 and high Ki-67. This raises the question, should we even be doing Ki-67 anymore? Should the FDA update its recommendation? It affects not just the United States but also other health authorities, which are influenced by its recommendations. I wanted to summarize those data quickly. I’m interested to hear, especially from Michelle. What do you think we should be doing, if anything, around Ki-67 given the challenges with testing?

Aditya Bardia, MD, MPH: Thanks for the nice summary, Heather. Michelle, your thoughts on Ki-67?

S. Michelle Shiller, DO, AP/CP, MGP: I continue to maintain that Ki-67 is informative. You see it reflected in Oncotype as well as part of its stratification of patients. Payers may not like to hear me say that, but it brings value beyond therapeutic decision-making, in terms of the baseline of what you’re dealing with and the proliferative nature of the tumor. I favor continuing to do Ki-67 beyond the treatment decision point. That’s my preference. We run it even though we have trouble getting paid for it.

Aditya Bardia, MD, MPH: How about you, Virginia? Are you a believer in using abemaciclib based on the FDA label? Or do you feel that should be revised and be more consistent with the ASCO guidelines?

Virginia Kaklamani, MD: I’m hoping it gets revised. Obviously, they were waiting for longer follow-up, but we now have 42 months of follow-up. We have a clinically significant absolute benefit in favor of abemaciclib for the whole group of patients. I use the criteria from monarchE, which are more in line with ASCO and NCCN guidelines. When you look at the Ki-67 low vs high groups, the absolute difference between abemaciclib and not was the same. What was different was that the patients with low Ki-67 had better outcomes than patients with high Ki-67, as you would expect.

Aditya Bardia, MD, MPH: Sara?

Sara A. Hurvitz, MD: I’m with my colleagues on this topic. If we have a high-risk patient, there isn’t clear evidence that their benefit differs based on tumor Ki-67. This requirement may have made sense early on, when we weren’t sure whether the curves were going to merge. You wanted to make sure it wasn’t going to be a PENELOPE-B situation, where at 40 months they’re exactly the same. From a health-economic perspective, that makes sense, and I can understand it. But my hope is that that will change now that we see this durable benefit.

Aditya Bardia, MD, MPH: Absolutely. That was the concern, and that’s why the additional follow-ups presented at SABCS were important. They emphasized that you continue to see a separation in the curves.

Transcript edited for clarity.