CDK4/6 Inhibitors Make Headway in HER2+ and Triple-Negative Breast Cancers

CDK4/6 inhibitors alone and/or in combination with HER2-directed therapies have been shown to elicit encouraging response rates and inhibit the growth of the cancer cells in HER2-positive breast cancer, according to Ruth O’Regan, MD. Although these agents are being combined with antiandrogens in triple-negative breast cancer (TNBC), more research is needed to determine which subtypes might benefit most from such an approach.

“The use of CDK4/6 inhibitors has been largely restricted to the estrogen receptor (ER)–positive and HER2-[negative] setting, but there are emerging data to suggest that HER2-positive cancers, particularly if they are hormone receptor (HR)–positive, may also benefit from [these] inhibitors, said O’Regan. “Also, in the space of androgen receptor (AR)–positive TNBC, some data suggest that [CDK4/6 inhibitors] may enhance antiandrogen therapy, although in unselected TNBCs, they don't appear to have much activity as single agents.”

In an interview with OncLive, O’Regan, chief of the Division of Hematology, Medical Oncology and Palliative Care within the Department of Medicine and associate director of Clinical Research at the University of Wisconsin Carbone Cancer Center, discussed the promise of CDK4/6 inhibitors in HER2-positive and triple-negative breast cancers and the need to identify predictive biomarkers for these agents to continue to propel progress forward.

OncLive: Could you speak to the rationale for examining CDK4/6 inhibitors in HER2-positive and triple-negative breast cancer?

O'Regan: CDK4/6 inhibitors have been shown to be highly effective in metastatic HR-positive, HER2-negative breast cancer, with really dramatic improvements in progression-free survival (PFS), and in some trials with overall survival (OS), when compared with endocrine therapy alone. However, less data are available on the use of these inhibitors in [those with] HER2-positive [disease], and particularly in those with TNBC.

One of the reasons for this is that, initially, when palbociclib (Ibrance) was being developed preclinically, investigators tested a whole bunch of cell lines to see whether it would be effective or not. Overall, they found that it was the ER-positive luminal cell lines that benefited from palbociclib. However, if you look at those data more closely, you can see that HER2-positive breast cancer cell lines, whether they're ER-negative or -positive, appear to benefit from palbociclib; however, TNBC appears to be more resistant in the initial studies that were done.

Following this, many preclinical studies have examined the use of CDK4/6 inhibitors in HER2-positive breast cancer, both in those with ER-positive and ER-negative [disease]. There does appear to be some efficacy with these single agents and also with HER2-directed agents in preclinical models. Even in preclinical models of cancer that is resistant to trastuzumab (Herceptin), if you give CDK4/6 inhibitors with HER2-directed agents, particularly if you add in chemotherapy, it appears that they are very effective in inhibiting the growth of those cancers.

Could you expand on some of the data that are available with these agents, specifically in the HER2-positive setting?

On the clinical front, the largest trial that's been done so far in the HER2-positive setting is the MonarcHER study. This study basically took patients with HER2-positive metastatic breast cancer who received at least 2 prior lines of treatment, that had to contain a taxane and ado-trastuzumab emtansine (T-DM1; Kadcyla). All of the patients had HER2-positive and HR-positive breast cancers. They were randomized to receive a standard arm, which was trastuzumab and investigator's choice of chemotherapy, compared with either abemaciclib (Verzenio) plus trastuzumab or abemaciclib/trastuzumab with the addition of fulvestrant (Faslodex). Results from that trial showed a 2.6-month improved PFS in the triplet arm compared with the other 2 arms. The abemaciclib/trastuzumab arm was equivalent to the trastuzumab/chemotherapy arm [with regard to PFS]. The response rates in the triplet arm were actually around 30% compared with only about 14% in the other 2 arms.

As far as toxicity, unfortunately, the triplet arm was a little bit more toxic. More grade 3 toxicities and more serious adverse effects (AEs) [were reported] in the triplet arm. Overall, none of the regimens were particularly toxic, so this would be an option for patients with HER2-positive, HR-positive breast cancer, particularly if they progress on T-DM1. The caveat to that is, of course, we have so many agents—3 have already been either approved or nearly approved for HER2-positive breast cancer this year—so [we are] not quite sure where this would actually fit in.

I also want to discuss a preoperative, phase 2 study in patients with HER2-positive breast cancer who were candidates for a preoperative approach. These patients received trastuzumab plus pertuzumab (Perjeta), a CDK4/6 inhibitor, and fulvestrant. Those patients actually had a very high response rate. The overall response rate was almost 100% and the pathologic complete response rate was 27%, which is pretty consistent with what you see with this kind of dual HER2 blockade and ER blockade in this setting.

Other trials are ongoing in the HER2-positive breast cancer [space]. One example is a trial examining T-DM1 and palbociclib, where patients are randomized to either T-DM1 alone or T-DM1 plus palbociclib. Interestingly, that trial is open not just for HER2-positive, HR-positive cancers, but also for HER2-positive, HR-negative cancers. This trial is currently recruiting patients. Just within the HER2 field, many new agents are out there, so that may be why we haven't seen a CDK4/6 inhibitor approved in this setting yet.

Have any notable research efforts been made for triple-negative disease with these agents?

In the triple-negative setting, if you look at preclinical studies, the basal-like triple-negative cell lines appear to be pretty resistant to palbociclib. The initial phase 1 trial that was done in retinoblastoma tumor suppressor–positive cancers basically showed no activity in TNBC, so we put that on the back burner. However, when you look at TNBC, there is not just 1 type of breast cancer; there are 4 to 6 different types under that triple-negative umbrella. One of them is the luminal AR subtype, which looks like a luminal ER-positive subtype genomically, but is actually ER negative and AR positive. This led to some trials looking at single-agent antiandrogens to treat these AR TNBCs. Overall, both enzalutamide (Xtandi) and bicalutamide (Casodex) produced pretty modest activity. The question become, “How could we make these antiandrogen drugs more effective in these AR-positive TNBCs?”

Just thinking about the ER-positive setting, it makes sense that you would look at a CDK4/6 inhibitor. Some preclinical studies [were] first done in prostate cancer that showed CDK4/6 inhibitors are actually effective in enzalutamide-resistant prostate cancer. This was followed up by a couple of studies in AR-positive TNBC. Results showed that if you look at TNBCs, the subtype most sensitive to CDK4/6 inhibitors is actually AR positive. In fact, the efficacy of CDK4/6 inhibitors in these TNBCs is actually related to AR expression. That was quite encouraging, and further preclinical studies showed that adding CDK4/6 inhibitors with antiandrogens appeared to effectively inhibit the growth of these tumors.

This has recently been looked at in a clinical trial at Memorial Sloan Kettering Cancer Center (MSKCC) where they took patients with AR-positive breast cancers and treated them with palbociclib and bicalutamide. The results of that were presented during the 2020 ASCO Virtual Scientific Meeting in a poster discussion. Overall, they found that there was a clinical benefit at 6 months in about one-third of patients, so really some clear activity for that combination. Previously, it has been shown that patients need to have greater than 10% AR expression for these types of approaches to work. In truth, we don't really know which type of AR-positive TNBCs [will benefit from the] addition of antiandrogens [and] CDK4/6 inhibitors.

We're currently doing a trial that is very similar to the one I just mentioned. Again, it's for [patients with] AR-expressing TNBC with AR expression of greater than 10%. In this trial, patients receive bicalutamide with ribociclib (Kisqali). The difference with our trial is that patients get a run-in of 2 weeks of bicalutamide before starting ribociclib. The reason for that is we wanted to try and look at resistance mechanisms of antiandrogens in these TNBCs. We're taking [count of] circulating tumor cells before patients start treatment and then after the 2 weeks, we use bicalutamide to see whether we can get an idea of what the resistance mechanisms are. Then, patients will go on to receive ribociclib with bicalutamide until disease progression.

We're just completing the phase 1 part of that trial and we did not notice any dose-limiting toxicities. We hope to open the phase 2 study quite soon. Then, hopefully, we may actually [try to look at some of] the tissue from the MSKCC trial once our trial is completed to see whether we can work out which patients really benefit from this approach. It is a reasonably nontoxic approach for patients with TNBC and much more targeted than some of the other treatments that we use.

Has any headway been made with identifying predictive biomarkers for these agents?

There has been extensive work looking at biomarkers for CDK4/6 inhibitors. They've looked at all proteins involved in the retinoblastoma and CDK4/6 pathways, and nothing really has panned out as being definitive. In fact, the only real biomarker that has been identified so far is ER itself. However, looking at some of the studies that we discussed, it's certainly possible that there are cancers that are ER negative that may benefit from CDK inhibitors, but we'd have to do a pretty deep dive to work out which cancers.

There has been a huge amount of work, both at the tissue level and also at the liquid biopsy level, to try to work out biomarkers that would tell us which patients would respond to a CDK4/6 inhibitor. Of course, we have a couple of adjuvant trials that we haven't seen full results of. One study, which used palbociclib, was reported as being negative.; in other words, the addition of palbociclib didn't significantly improve outcomes. Then, in the trial with abemaciclib, investigators actually did see a signal and improved outcome with the CDK4/6 inhibitor. Those trials might provide us with some biomarker information. However, again, it may be different in the early-stage [versus] the metastatic setting. We are on a quest to find that biomarker, but we haven’t found it yet.

Are any prominent CDK4/6 combinations emerging?

In the ER-positive setting, there is interest in looking at inhibiting CDK4/6, ER, and the PI3K/mTOR pathway. Several studies are examining that approach, either at the mTOR or the PI3K level. The data so far suggest that there is efficacy with triplets; it may not be better than with just a CDK4/6 inhibitor and endocrine therapy. We don't know yet because we don’t have direct comparisons.

The use of triplets is associated with more toxicity, although, they're not really overlapping toxicities, so it is a feasible approach. But again, I have always been leery of that because the results with CDK4/6 inhibitors and endocrine therapy in the first-line setting are so impressive that I don't know if you want to add another drug in that actually may be effective later on. The mTOR/PI3K inhibitors probably work better when the cancer is endocrine-resistant, so it may not make sense to give them early on.

I would imagine some data suggest that CDK4/6 inhibitors can make ER-positive breast cancers more immunogenic. As such, some trials are combining these inhibitors with immunotherapies, as well.