Case 1: MonarchE Study in High-Risk, HR+ Early Breast Cancer
Heather McArthur, MD; Massimo Cristofanilli, MD, FACP; Sunil Badve, MD, FRCPath; and Joyce O’ Shaughnessy, MD, review the study design and clinical implications from the monarchE study for high-risk, HR+ early breast cancer.
Joyce O’ Shaughnessy, MD: Heather and Massimo, from a practical standpoint, what would you recommend for adjuvant chemotherapy and adjuvant endocrine therapy for this post menopause? Let’s assume she’s a postmenopausal patient given the constellation we just went over. Heather, do you want to tell us what you would generally recommend?
Heather McArthur, MD: Certainly. I still use a lot of dose-dense ACT [doxorubicin (Adriamycin), cyclophosphamide, paclitaxel (Taxol)] chemotherapy for my very high-risk patients. TC [docetaxel (Taxotere), cyclophosphamide] would also be a reasonable regimen. I’d recommend adjuvant chemotherapy for this patient based on her oncotype score of 33, node positivity, and Ki67 consistent with the RxPONDER data for the node-positive postmenopausal population. I’d recommend chemotherapy. And because she’s postmenopausal, the adjuvant endocrine therapy that I’d recommend would be an aromatase inhibitor [AI], probably with abemaciclib [Verzenio]. That’s given the nodal involvement and the Ki67 greater than 20% in alignment with the monarchE data.
Joyce O’ Shaughnessy, MD: Thanks. How about you, Massimo?
Massimo Cristofanilli, MD, FACP: I’d also recommend chemotherapy. This woman certainly was young postmenopausal, so probably dose-dense AC [doxorubicin, cyclophosphamide]. If she can tolerate it well, then it will be the best choice for her. In terms of endocrine therapy, we only had aromatase inhibitors until a year ago, so I wouldn’t feel comfortable only doing that. We’re going to discuss a little more about the data and the use of CDK4/6 abemaciclib in combination with endocrine therapy for these high-risk patients.
Joyce O’ Shaughnessy, MD: Thank you. I want to ask Sunil something. Heather brought up the issue of ACT, which is probably her choice, but you wouldn’t get heartburn if somebody gave the patient TC. That’s a big question: who needs the anthracycline? I’ll tell you, rightly or wrongly, that I associate a high Ki67 in ER [estrogen receptor]–positive disease with more homologous recombination deficiency, more DNA repair abnormalities, a 70-gene signature, the ultra-high, the high 2, or maybe more basal-like. The higher the Ki67, such as if it gets 30%, 40%, or 50%, I start thinking, “I’d better get a DNA-damaging agent in there.” With cyclophosphamide, but I’m also more apt to add the doxorubicin in there. On the other hand, if somebody has 3 nodes but they’re lower Ki67, I’m not as anxious about the doxorubicin. Do you think there’s an association with Ki67 and homologous recombination deficiency in breast cancer?
Sunil Badve, MD, FRCPath: There are anecdotal data, but I’m not quite sure if people have done a proper study to document that. Clearly, there’s a gut feeling that the high proliferation may be associated with DNA repair. And we know that chemotherapy primarily attacks dividing cells, so a very high Ki67 would definitely be an indication that this patient would do well with a DNA-damaging agent.
Joyce O’ Shaughnessy, MD: Yes, I completely agree with you guys. That’s good to hear. I know there aren’t hard and fast data, but your gestalt is similar to mine. We talked about some of the risk factors that we were already worried about, so I won’t focus much more on that. But I’d like to run through some of the newer monarchE data, the updated data from the October ESMO [European Society for Medical Oncology] Virtual Plenary so we’re all on the same page with the new data.
But first, I want to remind us of these data from the neoMONARCH trial, where patients were randomized to 2 weeks of preoperative anastrozole [Arimidex] in the orange or abemaciclib and anastrozole in the dark blue and just abemaciclib alone in the royal blue. The primary end point was geometric mean change in Ki67. You can see in the top middle there that there was 90-plus percent reduction in mean Ki67, with the abemaciclib in 63%, with the AI alone, and then complete cell cycle arrest, which was less than 2.7% Ki67 after the 2 weeks. It was a lot higher with the abemaciclib. We can see it as well in these waterfall plots.
It’s clearly a very strong anti-proliferative agent. With these kinds of data, we know the 2-week marker has been predictive in past studies of benefit from AIs over tamoxifen, potentially predictive of what we might see in the adjuvant setting. It isn’t only the case with abemaciclib and palbociclib [Ibrance], but also with ribociclib [Kisqali], all of the CDK4/6 inhibitors. In the green, you can see suppression of Ki67 with letrozole [Femara] alone. The other 3 curves are with the addition of palbociclib. We see the same picture, but with a much greater depth. The complete cell cycle arrest is 59% with letrozole and 90% with palbociclib added. That’s very strong, and we know that quite well.
That led to the monarchE adjuvant trial. Patients could have received neoadjuvant chemotherapy. Ninety-five percent got chemotherapy, either neoadjuvant or adjuvant. It was a very high-risk population, both anatomically and molecularly higher risk, with 4 or more lymph nodes positive. If patients had 1 to 3 nodes, they either had to have grade 3 disease or T3 or T4 disease, or essentially confirmed Ki67 20% or higher. Nobody who was negative was in the study. After they finished their surgery, radiation, and chemotherapy, they were randomized to endocrine therapy of physicians’ choice vs the 2 years of full dose of abemaciclib. The primary end point was invasive disease-free survival [IDFS].
The most recent analysis was done in April 2021, with 27 months median follow-up. That’s enough follow-up that we could look at the 3-year IDFS rates. We can see at that point that 90% of the patients were off study treatment and 72% had completed the 2-year period. That’s because 10% were still getting treatment, but some of the patients stopped prematurely, either because of toxicity—I think that was 16%—or because of patient choice. Not everybody made the 2 years, although that number will go up by the time we see the final data.
It was a super high-risk population, with 60% of patients having 4 or more lymph nodes positive, 95% having had chemotherapy, and 43% were premenopausal patients. Down on the bottom, you see very few grade 1s, but a mix of grade 2s and 3s. Central Ki67 was required in the cohort 2, which was the 1 to 3 nodes positive with a Ki67 of 20% or higher. The tissue was strongly recommended to be sent in in cohort 1, but it wasn’t obtained from everybody. As you see here, we don’t have the Ki67 results in 20% of patients, and you can see that of those we do—the 80%—the 33% total of the population were less than 20%, and about 45% had 20% or more Ki67. That’s how it splayed out there.
This is the primary end point, which is the intent to treat population. In the IDFS data with the 27-month median follow-up, we’re looking at a 3-year IDFS. We can see a delta of 5.4% absolute improvement in IDFS, from 83.4% with endocrine therapy alone up to 88.8% with the addition of the abemaciclib. The hazard ratio is 0.69, which is significant. That’s the intent-to-treat population. All subgroups essentially benefited. I won’t go through the details, but no clinically-relevant subgroup didn’t benefit. Grades 2 and 3 benefited equally, for example. The distant relapse-free survival rate was 4.2%, an absolute improvement. It was also highly statistically significant, 0.0001, and a hazard ratio of 0.68.
The data are early, so it was important to look at how the hazard ratios were progressing over time, because this is the third report from the monarchE trial. You can see the hazard ratios with both IDFS and distant relapse-free survival as we went from year 0 to 1, years 1 to 2, and years 2-plus. Ninety percent of the patients are finished with their abemaciclib, and we can see that the hazard ratios are continuing to get smaller. We need further follow-up, but at least so far so good as patients stop their abemaciclib and get dumped out from their treatment. The hazard ratios are still going in the right direction.
Looking at the overall high Ki67 population on the left, and the intent-to-treat—everybody that was known to be high Ki67—and then looking at cohort 1, which combined the high Ki67 with high anatomic risks—4 or more nodes, or grade 3, or T3 disease—we see that the greater the risk the patient has, the greater the absolute benefit in IDFS. For example, in cohort 1, high Ki67, what the FDA [Food and Drug Administration] decided to approve is on the right, and that was a 7.1% improvement in IDFS. I think the FDA thought, “That’s a somewhat big delta for this high-risk group that’s primary endocrine therapy resistant, because they’re recurring so quickly on endocrine therapy.” I think they felt like this was a high-risk group, and they just went ahead and approved it even though the median follow-up is still quite short.
But our colleagues in ASCO [American Society of Clinical Oncology] and NCCN [National Comprehensive Cancer Network] took a look at this curve and said that there’s a big improvement in outcome. The solid curves make up the high Ki67 population for IDFS, but the dotted lines are the low Ki67 patients, and they also benefited. It’s a smaller number of events, but they also benefitted. The hazard ratios weren’t that different between the high and low Ki67. If you have a lobular cancer that isn’t a high Ki67—but she’s got 8 positive nodes—our ASCO and NCCN guideline colleagues said she’s a reasonable candidate because she’s part of the intent-to-treat population in monarchE.
I wanted to show survival here as well because this played into why the FDA chose the cohort 1 high Ki67 population. This is the FDA-indicated population. These survival data are extremely early. You can see the number of events is less than 100. But you see the hazard for survival is 0.767. There wasn’t a trend in the intent-to-treat population. There isn’t a trend at all in survival, but for this FDA-indicated population, there’s a trend emerging a little in survival. That played a part in why the FDA said, “Let’s focus on this for the earliest of indications here.”
As far as safety, there’s mainly diarrhea and tachyphylaxis. We want to get some opinions from you guys about what you’ve seen. But generally speaking, with dietary modification, it’s quite manageable. It tends to get better after 4 to 6 weeks. Venous thromboembolism [VTE] was 2.5%. And if you look down at the bottom box, any grade VTE was 4.1% with tamoxifen and 1.7% with the AI. The risk of ILD [interstitial lung disease] is 3.2% with abemaciclib and endocrine therapy, but it tends to be lower grade. There weren’t any deaths or anything serious.
Transcript Edited for Clarity
Breast Cancer Leans into the Decade of Antibody-Drug Conjugates, Experts Discuss
September 25th 2020In season 1, episode 3 of Targeted Talks, the importance of precision medicine in breast cancer, and how that vitally differs in community oncology compared with academic settings, is the topic of discussion.
Listen
