Case Discussion: Resistance to CDK4/6 Inhibitors in HR+ mBC

Andrew D. Seidman, MD: To put a human face on this, I’d like to present the case of a 56-year-old postmenopausal woman who was diagnosed with stage IIB, hormone receptor–positive, HER2 [human epidermal growth factor receptor 2]–negative breast cancer on core biopsy. She originally underwent breast-conserving surgery. We were told that she had node-negative disease and a recurrence score of 28. She received adjuvant docetaxel and cyclophosphamide, had adjuvant radiotherapy, and was placed on an adjuvant aromatase inhibitor [AI]. Four years into her adjuvant aromatase inhibitor, she presents with coughing and some dyspnea. Unfortunately, imaging shows metastatic recurrence involving lung as well as bone in her thoracic spine and ribs. This is confirmed by lung biopsy as an ER [estrogen receptor]–positive, HER2- breast cancer, similar to her primary tumor.

This occurred while on adjuvant AI. Then she received fulvestrant plus ribociclib as her first therapy for metastatic disease. She also had stereotactic radiotherapy to the spine. She did well with this for almost 2½ years, but now she’s progressed further in the same organ systems in lung and bones. Her ECOG Performance Status is 1, her laboratory abnormalities are normal, and the therascreen PCR [polymerase chain reaction] assay was sent for PIK3CA mutation and revealed the E545K hot spot mutation. Would anyone like to begin the discussion on this 1? Hope, do you want to tell us how you would think about this patient who enjoyed 2½ years of progression-free survival [PFS] on first-line fulvestrant plus ribociclib but has now progressed and has a hot spot PIC3CA mutation?

Hope S. Rugo, MD: These are such tough situations when a patient relapses on their adjuvant endocrine therapy. I wish we had a better idea of how to recognize these patients up front and intervene earlier. Maybe some of the data we’ll talk about a little later will give us that path. In a patient like this, you’ve already used your SERD [selective estrogen receptor degrader]. You’ve given fulvestrant, so you can’t use SOLAR-1 as your primary direction. I usually don’t like to continue the same endocrine agent when a patient has developed progressive disease and has visceral metastases, although I’m sure we could all find exceptions to that.

In this situation, I would use the data from the BYLieve study, which is a nonrandomized phase 2/3 cohort trial. Cohort B, which was presented at 2020 San Antonio Breast Cancer Symposium, looked at the patient population who had progressed most in their last therapy on fulvestrant and a CDK4/6 inhibitor. They received the AI letrozole with alpelisib if they had a PIK3CA mutation.

What was important was that we looked back to see how many of those patients had received a nonsteroidal AI or any AI within a 6-month period of time of developing metastatic or progressive disease, and it was a very high number: in the 80% range. In those patients, we did meet our end point. It looks like the benefit in these patients who are more heavily pretreated might be a little less. But these are single-arm phase 2 trials compared with SOLAR-1, where patients had relatively less treatment. I’d use that data as a way to direct her treatment, and I’d give a nonsteroidal AI with alpelisib as her next treatment.

Andrew D. Seidman, MD: Thank you for that. It’s great to have data. The median PFS in that cohort B was in the range of about 6 months, which was better than expected.

Hope S. Rugo, MD: Yes, it was. We’ve had some discussions about that. Could it be alpelisib alone? We certainly see activity alone, but there may be some synergy that we’re not able to identify in a single-arm trial in which you give an AI that we wouldn’t normally give after a patient progresses. You add a targeted agent, and we’ve all seen that anecdotally where patients did much better than we would have expected when we take an approach like that.

Andrew D. Seidman, MD: There’s also been some recent real-world data to help us understand the clinical trial data. Do you find that data to be helpful at all?

Aditya Bardia, MD, MPH: Yes. It’s helpful more from a safety tolerability perspective. I would always exert caution in interpreting the PFS. In these studies, it’s not the same trigger. You don’t have scans at the same frequency, so just looking at the PFS could be problematic. But from a safety tolerability, the discontinuation is very helpful.

Coming back to what was discussed earlier about real-world studies at the 2021 ASCO [American Society of Clinical Oncology] Annual Meeting, we had the experience from the French group, which looked at their expanded access program. Their experience was that the combination was well tolerated and had efficacy. They did some multivariate analysis to see if there are certain predictors in terms of who would really do well with fulvestrant plus alpelisib. They found that prior use of everolimus was not a predictive factor in terms of doing worse or better with alpelisib. Surprisingly, prior use of fulvestrant was associated with improved outcomes with fulvestrant plus alpelisib.

Taking the caveats of a real-world study, you could consider fulvestrant plus alpelisib for this patient. Although I agree with Hope that, in general, we like to switch the endocrine partner. When someone has progression on letrozole plus CDK4/6, we’d like to do fulvestrant plus something else. It’s just the opposite here. With a patient who had progression on fulvestrant plus ribociclib, I’d like to do an AI plus alpelisib. I’d also like to get plasma-based genotyping in this setting. We saw that there was a PIK3CA mutation, but it’s important to ensure that there’s no ESR1 mutation. If this patient had an ESR1 mutation, I would not use an AI and would continue fulvestrant.

Hope S. Rugo, MD: I still would if there was an ESR1 mutation. There are more of those differences in the more resistant ESR1 mutations, as Dejan was talking about earlier. We didn’t look at this in the population of patients. We know there was a percentage of patients in BYLieve who had ESR1 mutations. That’s what we see. We see them acquired with fulvestrant as well as AI. I wouldn’t use that as a differentiating factor until we know for sure that’s not a benefit. We also had a poster at ASCO this year with Foundation Medicine in which we looked at all the mutations that were not SOLAR-1 mutations. SOLAR-1 used the specific, more frequent mutations—I asked them, “How did you select these?” Dejan can comment on that, but there was actually no answer. It was just about choosing the more frequent ones. The interesting thing is that in the real-world setting, we seem to see benefit across this whole spectrum of the PIK3CA mutations that we saw.

Andrew D. Seidman, MD: Clearly the resistance associated with ESR1 mutations isn’t absolute but rather relative. There’s some wiggle room.

Transcript Edited for Clarity