Biomarkers Could Be Key in Bringing CDK4/6 Inhibitors to Adjuvant HR+/HER2- Breast Cancer

Keerthi Gogineni, MD, MSHP

Conflicting results regarding the benefit of adjuvant CDK4/6 inhibition in early-stage hormone receptor (HR)–positive, HER2-negative breast cancer have prompted many questions regarding patient selection and treatment duration, which many hope will be answered with ongoing biomarker research, explained Keerthi Gogineni, MD, MSHP.

“Some of the studies that are ongoing right now and are being developed [are looking at] women who are on endocrine therapy [to see if there is] a way that we can identify risk of relapse through circulating tumor cells,” said Gogineni. “If [the patient’s risk] crosses a certain threshold, is that a person in whom then we add therapy or augment their endocrine therapy? Perhaps that would be a way for us to identify a higher-risk group of patients that could benefit [from intensified treatment]. However, this is all to be seen.”

In an interview with OncLive® during an Institutional Perspectives in Cancer webinar on Breast Cancer, Gogineni, an assistant professor in the Department of Hematology and Medical Oncology at Winship Cancer Institute, Emory University School of Medicine, discussed the data with adjuvant CDK4/6 inhibition in early-stage, HR-positive, HER2-negative breast cancer and biomarker research that could help inform treatment recommendations in the future.

OncLive®:Given the data we have with CDK4/6 inhibitors in the metastatic setting, was there an expectation that a consistent benefit would be seen in the adjuvant setting?

Gogineni: Very much. We had been so pleased with the outcomes, both in terms of progression-free survival [PFS] and now even overall survival [OS] in the advanced setting. The natural extension of that was to see if [CDK4/6 inhibitors] reduce the risk of recurrence for women who have higher-risk, early-stage disease. Unfortunately, 3 trials were [presented] this year regarding CDK inhibition in the early-stage setting and 2 of the studies that [evaluated] palbociclib [Ibrance], even though they had enrolled a high-risk population, didn’t show an added benefit to using palbociclib alongside endocrine therapy at 2 years. There may have been various reasons for that.

In PENELOPE-B, the duration of palbociclib may have been too short, at 1 year. In PALLAS, there were high discontinuation rates in the treatment arm. We may not have seen the benefit of palbociclib because we didn’t have as many patients stay on drug. The other issue, potentially, was with respect to risk. In comparison with monarchE, which evaluated abemaciclib [(Verzenio) in the adjuvant setting], [patients in PALLAS] were less high risk than the patients that were enrolled in monarchE. It may be that we need to really carefully select the population for whom this is going be a good strategy. Even in monarchE, although we saw a statistically significant benefit at 2 years, there’s questions about whether the curves are going to cross with longer follow-up. We’ll have to stay tuned to see whether or not that’s the case.

A third study is evaluating ribociclib [Kisqali] in this setting, and we await those results, but I believe it’s going to be some time before we get to see some data from that trial.

Have we been able to tease out biomarkers of response to CDK4/6 inhibitors?

One of the biomarker studies that was [presented at the 2020 San Antonio Breast Cancer Symposium (SABCS)] was regarding the MONALEESA studies. When investigators looked at the specimens from women who had been on ribociclib, and they did intrinsic subtype analysis on these specimens, they could see an association between those intrinsic subtypes and the likelihood of benefiting from ribociclib in that metastatic setting.

Not surprisingly, most of the patients who were evaluated had luminal A–type breast cancer, which is what we would expect. Interestingly, even though these were all estrogen-positive, breast cancer specimens, we did see some HER2-enriched subtypes as well as basal-like subtypes. Interestingly, the HER2-enriched [population] seemed to benefit the most [from ribociclib], whereas the basal subtype didn’t benefit at all. When we move towards doing more intrinsic analysis on patients, maybe we identify the patients who are less likely to respond [to a CDK4/6 inhibitor] and choose a different strategy for them.

The other interesting thing that we learned from monarchE was the results of patients who had high Ki-67 scores on their specimens. That group seemed to clearly benefit from that adjuvant CDK4/6 strategy, whereas the group that had the lower Ki-67 [scores] didn’t garner as much [benefit], so maybe this will also be a way for us to identify which group is going to benefit the most from an adapted adjuvant strategy.

Was the Ki-67 cutoff chosen arbitrarily, or is there rationale to suggest that patients with a Ki-67 index below 20% would not benefit from CDK4/6 inhibition?

Ki-67 evaluation can vary from institution to institution, and the truth is, it’s not an automatic or standard part of pathology reports. It really depends on the center. The [investigators] had selected the cutoff to be 20%, but I do believe they’re going to be looking at whether other cutoffs have the same ability to predict outcome benefit and whether we need to set a higher or lower threshold. It does become harder to distinguish variation as that number gets lower. That might be part of the issue for us later.

What agents are under study in combination with CDK4/6 inhibitors in the metastatic setting to help extend their utility beyond progression?

One of the questions that we are trying to answer now is: What role may CDK4/6 inhibition have beyond progression? We routinely use these drugs in the first-line metastatic setting. Patients who had received other approaches in the first-line setting would certainly be offered [a CDK4/6 inhibitor] in the second-line or beyond setting. The question will be: If you progress on a CDK4/6 inhibitor, can you continue the CDK4/6 inhibitor but switch the endocrine partner? Or, can you augment [the response] by adding a third drug, whether that be a checkpoint inhibitor or a PI3K inhibitor. Those are questions and studies that are going on now to try to see if we can extend the benefit [of CDK4/6 inhibitors]. Combination therapy could be something that would be potentially a first-line approach for someone who had higher burden disease or had molecular testing that reflected a PIK3CA mutation. Right now, [combination approaches are] really being looked at in the second-line and beyond settings.

If CDK4/6-based triplets are ultimately brought into the adjuvant setting, but we’re already seeing adherence issues with adjuvant palbociclib plus endocrine therapy, how do you anticipate that will be addressed?

That is a really great question. Even with just endocrine therapy alone, compliance and adherence drops off after 5 years. When we’re asking patients to do some of these drugs for 10 years, what’s the likelihood that they’re going to be on it for that long? We saw that that was in fact an issue in PALLAS in terms of the drop off rate. PENELOPE-B had better adherence, as did monarchE.

We might want to ask a different question, which is: When [should] we initiate the CDK4/6 inhibitor? Late relapse is a concern in ER-positive, early-stage disease. More than half of relapses happen in years 5 to 10. Perhaps the strategy here is for those higher-risk patients who are thinking about initiation of the CDK4/6 inhibitor later. Also, we have to think about the duration [of treatment because] this is an expensive drug, and it’s not without toxicity. We really need to identify the right group of patients that’s going to benefit from this strategy.

Is circulating tumor DNA a robust enough tool, or are there still adjustments that need to be made before it is incorporated into practice?

The role for circulating tumor cells and trying to identify what is a threshold that should be something that raises alarm bells [is of interest], but also, we don’t know yet that intervening at that point actually changes outcome. Circulating tumor and cell-free DNA is also being used all the time as sort of a liquid biopsy mechanism for patients with metastatic disease to try to identify ESR1 and PIK3CA mutations.

An interesting bit of data came out of the SOLAR-1 update, showing that liquid biopsy was pretty good at identifying PIK3CA mutations, but did miss it in about 35% of people who then had PIK3CA mutations identified on tissue. We have to remember that you may see different things [with tissue and liquid biopsy]. The beauty of the liquid biopsy is that you’re more likely to look at real-time information. It’s hard to ask patients to get tissue biopsies over and over.

This might be a mechanism to sort of reduce your invasive procedure on a patient, remembering that if you have an opportunity to use a targeted drug, and you don’t see the results of the profile that you might be expecting on a liquid biopsy, to revisit the possibility of repeating that test on tissue because you might see a group of patients that could benefit from a drug that we would have otherwise not offered.

Shifting to PI3K inhibitors in the metastatic setting, can alpelisib (Piqray) be used with any endocrine therapy?

Generally speaking, we look to PI3K inhibitors in a later line of therapy because we still don’t see the PFS benefit [with that class of agents vs] what we see from first-line CDK4/6 inhibitors, but also, there’s a different toxicity profile that makes it a little bit of a harder drug to take.

The SOLAR-1 trial demonstrated benefit of alpelisib with fulvestrant [Faslodex] in patients with a PIK3CA mutation. However, a lot of patients may have had fulvestrant already. The question arose: Can we pair alpelisib with a different endocrine partner? The BYLieve trial reported out 1 of its cohorts at the 2020 SABCS, which looked at patients who had gotten a CDK4/6 inhibitor and fulvestrant previously, and then received alpelisib plus an aromatase inhibitor [AI. In that scenario], we still saw a PFS benefit. It was reassuring that, for those patients, we could still use a PI3K inhibitor and use it with a different endocrine partner that a patient hadn’t been exposed to yet.

How would you summarize the data with CDK4/6 and PI3K inhibitors in HR-positive, HER2-negative breast cancer?

We see incomparable PFS benefit with CDK4/6 inhibitors and endocrine therapy in the first-line setting. We also now have data from MONALEESA-7, that the CDK4/6 inhibitor with endocrine therapy has an OS benefit for premenopausal women.

In the early-stage setting, we saw some [mediocre] results for adjuvant strategies, depending on the CDK4/6 inhibitor, albeit promising data with abemaciclib in combination with endocrine therapy. However, we probably need to watch those data for a few more years to see if that benefit persists.

Additionally, knowing that we can pair an AI with a PI3K inhibitor in a patient who has the mutation but who already was exposed to fulvestrant [is encouraging]. Future directions that we look forward to seeing are the results of NATALEE, which will show us what 3 years of ribociclib looks like in the adjuvant setting alongside of endocrine therapy. [We also hope to answer whether] some of these biomarker studies point towards how we can better select patients for these strategies.