How Do You Decide on an Appropriate Endocrine Therapy?

Debu Tripathy, MD:The exclusion and inclusion criteria used in the CDK4/6 inhibitors have been rather uniform. There have been a few exceptions here and there, but for the most part, they called for hormone receptor—positive breast cancer using CLIA guidelines. They used NCCN guidelines to define menopausal or premenopausal, depending on the trial. They included patients with bone-only disease if they had at least 2 lesions and 1 that was measurable, a lytic lesion. They also required that patients had acceptable liver function tests. They could have some abnormalities in their transaminases, but there were some limitations in that area. They had to not have significant prolongation of the QTc interval to start with. Those were generally the eligibility criteria.

There are many factors that we have to consider when treating patients with hormonal therapy or hormonal and biotherapy. One of them is the tempo of the disease, and that can be measured in several ways. One way is, what is the disease-free interval from when they completed or had their adjuvant therapy for early stage disease to when they recurred? There’s a sense that there may be some relative resistance to hormonal therapy in patients who have recurred within 1 or 2 years of completing their adjuvant hormonal therapy, whereas patients who are de novo who present with metastatic disease seemed to be more sensitive.

Nevertheless, when one looks at the benefit of CDK4/6 inhibitors across these categories, they’re roughly the same. It’s about a doubling of the progression-free survival. Now, the absolute progression-free survival may be shorter in someone who has completed their adjuvant hormonal therapy more recently compared with someone who is de novo, but the benefit is still the same. We haven’t yet identified a clear group of patients who don’t benefit from the addition of CDK4/6 inhibitors.

I think that the kind of patient who might be a candidate for chemotherapy up front has very heavy visceral disease or is rapidly progressing or highly symptomatic. Those patients would not have been eligible for these trials, nor would we generally want to treat them with endocrine therapy anyway. I should also add that some of the rarer situations that were excluded in these trials have to be taken into account. Patients with brain metastases were excluded, although I generally don’t think that’s an exclusion for anything as long as they’re treated and stable. Patients with inflammatory breast cancer were excluded also. One might argue that’s a more aggressive disease that should be treated with chemotherapy, but it’s also a disease that’s very hard to follow clinically, so judging responses would have been difficult on the trial. I think in carefully selected patients of either of those categories, one could use endocrine or endocrine biotherapy.

The second part of this issue is, who might be a candidate for endocrine therapy only? None of us can really identify who that patient might be. We just haven’t come across factors. There has been some recent description of certain factors like loss of retinoblastoma, which may occur on treatment or at baseline, or high cyclin E levels, which might be an escape through CDK2. But we need a little more validation of those to see if they might influence our decisions.

The primary and secondary resistance mechanisms of CDK action are still illusive. Initially, there was an attempt to look at retinoblastoma status, which is uncommon in hormone receptor—positive cancers; 80% to 90% of patients have intact RB. But the presence of cyclin D amplification, which is also seen in maybe a third of these patients, didn’t seem to have an impact.

Recently, as we’ve gotten more data from these trials and started to more comprehensively analyze baseline tissue and conduct some very elegant studies, where they were looking at multiple samplings along the way, while patients were on treatment using liquid biopsies and blood-based techniques, we have learned a little bit about some potential mechanisms. One of these is the loss of the retinoblastoma protein, which may be there at the beginning, but on serial blood draws and liquid biopsies when you look at the retinoblastoma gene, some patients do lose this.

Another interesting feature that has come up is amplification of the fibroblast growth factor receptor,FGFR. These seem to also come up serially, and there are actually some strategies now to targetFGFR. However, it is too early for us to know what the relative risk is. We’re not quite ready to use these in the clinic. But I think as time goes on and we validate these changes, we will be able to use them in the clinic, especially as we start routinely doing liquid and solid tumor biopsies.

One other common mutation, theESR1mutation—that is, the estrogen receptor alpha—seems to be associated with resistance to aromatase inhibitors. But interestingly in the PALOMA-3 study, when they looked atESRmutations, it didn’t seem to make a difference in prognosis or in the response to palbociclib added to fulvestrant. So, while we think that’s an interesting biomarker, we aren’t really using that in decision making yet.

Transcript edited for clarity.