Sequencing Therapy in HR+/HER2- MBC

Joyce O’Shaughnessy, MD: Let’s turn to how we sequence therapies in the metastatic setting. We said we’re almost always starting with the CDK4/6 inhibitor. That’s the easy part, but then it comes after that. With the PIK3CA mutation, in an appropriate patient who doesn’t have uncontrolled diabetes, we’re probably going to, when we can, alpelisib. Sometimes patients need chemotherapy. It’s a minority, but we’re utilizing that. There are a little data, not much, for CDK that’s getting traction in the community, although we’re light on the data. Yet some folks are sequencing the CDKs or coming back to 1 later on.

Let me start with you, Neil, if you don’t mind. Generally speaking, what’s the big picture? How do you think about sequencing the various subsets of patients after CDK?

Neil M. Iyengar, MD: I agree that sequencing therapy after exposure to CDK4/6 inhibitor is certainly an evolving landscape. We do have some data to help guide us. A lot of those data are extrapolated, but we could look back a couple of years ago to the MAINTAIN trial that was presented at ASCO [American Society of Clinical Oncology Annual Meeting]. That was a small study that was a little over 50 patients or so. That was our first clue that perhaps there’s activity of continuing a CDK4/6 inhibitor after prior exposure.

They used abemaciclib specifically in the post-CDK4/6 exposure setting. I know that a lot of my colleagues in the community are doing this. It’s a patient-by-patient decision. I’ve thought about and have done this in patients who perhaps have a minor progression of disease and aren’t dealing with large-volume disease, or they have a mutational profile. The data are evolving and mutational profile may not be consistent with known resistance mutations or mechanisms. But in terms of strong evidence, we’re not there.

In the post CDK4/6 space, moving on to alpelisib in someone who has a PI3 kinase mutation is warranted now that we have some data from BYLieve. But it’s an evolving space. For those who don’t have a PI3 kinase mutation, that’s where we’re struggling with that decision-making. Everolimus and exemestane have been extrapolated from BOLERO-2 as a potential use, but biologically, mechanistically speaking, this may not be an efficacious approach. We do have some data from the EVERMET study, where the use of everolimus mTOR inhibition after prior CDK4/6 inhibition wasn’t particularly effective. I’ve done this occasionally in patients with minor progression of disease, but I’m doing it less and reaching for a clinical trial in this space.

Joyce O’Shaughnessy, MD: Thank you, Neil. Sarah, how do you think through what to do after CDK4/6 progression?

Sarah Sammons, MD: That’s where our next-generation sequencing comes in handy. If the patient has a PI3 kinase mutation and they’re fit, I’m reaching for alpelisib with an endocrine therapy backbone. If they don’t have a PI3 kinase mutation, and I don’t have a clinical trial for them to go on, if they’re fit, I still reach for everolimus. I haven’t reviewed the data that Neil just spoke of but, we have data with everolimus with the fulvestrant backbone. We have data with everolimus with the tamoxifen backbone. I’ve had patients who have meaningful clinical benefit on everolimus combinations after CDK4/6. So I’m looking for a clinical trial, looking at their next-generation sequencing, and everolimus combination.

Joyce O’Shaughnessy, MD: Do you ever use or go back to a second CDK? Have you used palbociclib or ribociclib? Would you use abemaciclib?

Sarah Sammons, MD: We were just talking about that today. The most data that we have is mostly retrospective. There are ongoing prospective studies. But the data that we have suggests that abemaciclib might be fairly effective after progression on palbociclib or ribociclib. In my practice in the last week, I had a patient who very clearly had developed 2 new bone metastases but no visceral disease. We changed the endocrine therapy backbone, and we changed the palbociclib to abemaciclib to see if we could get some longer benefit out of his CDK4/6 combination. I’ve done that, and data are evolving in certain circumstances.

Joyce O’Shaughnessy, MD: I was part of that retrospective analysis. It was a case series from 4 or 5 institutions. It’s data light, no question. The clinical benefit rate was about 36%. Some of the patients had quite a bit of time between the CDKs. Mostly it was palbociclib first line and then later on abemaciclib, or abemaciclib immediately thereafter. But a subset appears to benefit, and I’ve seen that in my practice.

Gabe, how do you approach the choices we have after CDK4/6?

Gabriel Hortobagyi, MD, MACP, FASCO: This is a relatively data-free zone, so it’s a rapidly evolving area. My first choice—I’m fortunate enough to live in a situation where I have several clinical trials to choose from at all times, so that’s my go-to. It’s also a complex discussion with the patient because this is 1 area where I might choose endocrine therapy alone. I might choose everolimus and exemestane or some other endocrine backbone. Of course, in the PI3 kinase mutant group, I will also discuss alpelisib. Different patients will make different choices in this situation. The standard algorithm would be mutant alpelisib, non-mutant everolimus, then subsequent treatment. But it’s more complicated than that. Many patients fall out because of comorbid conditions, many patients fall out because the benefit you offer them isn’t attractive enough for them. It’s a complex but important discussion with the patient because these drugs are clearly associated with greater toxicity, and the patient needs to understand that. He needs to understand that there will be more work required on her or his part and on my part to make this work safely and well.

Joyce O’Shaughnessy, MD: That’s a good groundwork with pluses and minuses. It’s going to be a little more effort on our part, but it’s a good option. It’s not chemotherapy. I’ll tell you my experience with everolimus as I’m progressing on CDK. For the bone-only patients and those with ER-dependent sites of disease, I get benefit. The bone-only patients will go on and get 6, 9 months, sometimes a year out of everolimus. If they’d benefit more than 6 months, then I can come in with single-agent tamoxifen. They escaped through the estrogen receptor. But I’ve found conversely that, if patients are progressing in their liver on CDK4/6, I’ve not had benefit with everolimus. I wonder if the EVERMET was skewed toward not showing benefit because there were some liver patients in there. This is just my experience. If I pick the bone-only patients, or the pleural disease, the pulmonary disease, or the lymph node disease, I get good results with everolimus if they don’t have a PIK3CA mutation. We’re figuring it out until we get more data.

Transcript edited for clarity.