Mature Data Drive New Insights With Novel Agents in Breast Cancer

The data with CDK4/6 inhibitors, PI3K inhibitors, antibody-drug conjugates (ADCs), and PARP inhibitors in metastatic breast cancer continue to support their use in the clinic, explained Lisa A. Carey, MD, FASCO. However, to date, only PARP inhibitors have shown sufficient evidence that warrant their movement into the adjuvant setting.

In interview with OncLive® during an Institutional Perspectives in Cancer webinar on breast cancer, Carey, who chaired the event, and is The Richardson and Marilyn Jacobs Preyer Distinguished Professor in Breast Cancer Research and Deputy Director of Clinical Sciences at Lineberger Comprehensive Cancer Center, UNC Chapel Hill, discussed emerging and updated data with CDK4/6 inhibitors, PARP inhibitors, ADCs, and PI3K inhibitors in breast cancer.

The virtual meeting covered optimal treatment selection in HER2-positive breast cancer, changing standards for hormone receptor (HR)–positive breast cancer, encouraging data in triple-negative breast cancer, and practical applications of genomic assays in estrogen receptor (ER)–positive breast cancer.

OncLive®: Did the updated analysis of the phase 3 PALOMA-3 trial (NCT01942135) confirm the benefit of adding palbociclib (Ibrance) to fulvestrant (Faslodex) in patients with advanced HR-positive, HER2-negative breast cancer? 

Carey: At the 2021 ASCO Annual Meeting, we saw updated overall survival [OS] data from the PALOMA-3 trial, which is testing palbociclib plus fulvestrant in patients with pretreated HR-positive, HER2-negative metastatic breast cancer. The [updated data] confirmed about a 7-month difference in favor of adding the CDK4/6 inhibitor to fulvestrant in patients in the second-line setting. [The combination] seemed to work well in most patients, maybe a little less so in patients with a short disease-free interval [DFI] before their cancer metastasized and those who had prior chemotherapy. Outside of that resistant population, it looked like [the combination] worked very well and across biomarker subsets, such as ESR1.

Were similar findings seen in the updated analysis of the phase 3 MONALEESA-3 trial (NCT02422615)?

The other important update came from MONALEESA-3, which is a similar trial. The study didn’t require pretreatment, so [about half of patients received] first-line endocrine therapy, [including] fulvestrant, with or without a CDK4/6 inhibitor. In this case, ribociclib [Kisqali].

The investigators reported and updated the OS end point with about a 12-month difference in median OS. We saw a nice augmentation in outcomes with ribociclib added to fulvestrant. Now, interestingly, in the first-line setting, which were patients who were either de novo or those with a longer DFI, did pretty well. That indicates that [palbociclib and ribociclib] work; they both demonstrated an OS advantage when added to fulvestrant.

In truth, the first-line subset in the MONALEESA-3 trial in terms of outcomes looks like the first-line trials with AIs [aromatase inhibitors]. I don’t know how relevant that is for the US population because most of the time we give the CDK4/6 inhibitor in the first-line setting, but [CDK4/6 inhibitors] clearly work with fulvestrant, as with anything. One of the take-home messages is that CDK4/6 inhibitors work regardless of the endocrine backbone.

Would it be preemptive to say whether CDK4/6 inhibitors will move into the adjuvant setting based on the data we’ve seen from the monarchE (NCT02107703), PALLAS (NCT02513394), and PENELOPE B (NCT01864746) trials?

monarchE is an adjuvant trial testing abemaciclib [Verzenio] in HR-positive, HER2-negative early breast cancer with certain risk features. The monarchE study was a big study of nearly 6000 patients. Patients had to have either 4 or more lymph nodes, or if they had 1 to 3 lymph nodes, they had to have other negative [prognostic] features, such as a [large] tumor [size], a high Ki-67 index, or high-grade disease. [The trial enrolled] node-positive patients with varying risk features, but really [encompassed] those with an escalated risk.

Patients were randomized to receive abemaciclib [or placebo] for 2 years. At the 2020 San Antonio Breast Cancer Symposium, the primary analysis, which was [done with] 19 months of follow-up, was presented. The investigators presented the 2-year invasive disease-free survival [iDFS] data that was in favor of adding abemaciclib in the adjuvant setting by about 3%. [The iDFS rate] was 89% [with alpelisib] vs about 92% [with placebo]. The investigators also reported the distant disease-free survival, which is a more relevant end point, and there was also about a 3% difference in favor of abemaciclib. The problem with that analysis is that at 19 months you’re reporting 2-year follow-up data, so it is an immature analysis and whether that holds up with more mature data remains unknown.

What was reported at the 2021 ASCO Annual Meeting was a prespecified subset of patients who had received neoadjuvant chemotherapy, and [that population comprised] about a third of the [overall population]. In those patients, there was a bigger difference in 2-year iDFS of 80% [with placebo] vs 87% [with abemaciclib]. This is a high-risk population, a little bit higher risk than the parent trial, and maybe [gives] a little bit more of a sense of abemaciclib impact, but the truth is, [these data] remain immature, and we’re all waiting to see how [these data] look once they become more mature.

The 2 negative trials with palbociclib in the adjuvant setting––PENELOPE B and PALLAS––were reported within the past year or so. A lot is riding currently on abemaciclib regarding the adjuvant use of [CDK4/6 inhibitors]. We’re awaiting more mature analysis, and we’re also awaiting ribociclib data. The jury is still out.

If in fact the monarchE data hold up, you’re moving abemaciclib into the adjuvant setting, which has implications for the first-line and later-line metastatic settings. Then you have to rejigger everything because currently we use CDK4/6 inhibitors pretty conventionally in first-line metastatic disease. If a patient has already been exposed [to a CDK4/6 inhibitor in the adjuvant setting], then we have to identify and address whether that therapy works in the first-line metastatic setting.

Speaking of adjuvant approaches, might PARP inhibitors become a standard of care in the adjuvant setting for patients with HER2-negative disease based on findings from the OlympiA trial (NCT02032823)?

OlympiA is a practice-changing study. The extent of all the other trials that we talked about have practice implications, some of them are practice clarifying, but OlympiA is really practice changing. This was a study of just under 2000 patients with germline BRCA1/2 breast cancers. They were all HER2 negative, and they all had to meet criteria for being relatively high risk clinically, and they all had gotten chemotherapy. Patients were randomized to receive olaparib [Lynparza] vs [placebo], and the investigators showed a stunning impact with the addition of olaparib; we saw a 3-year iDFS rate of 77% in the control arm, and it went up to 86% [with olaparib], with a hazard ratio [translating to] nearly a 40% to 50% difference in favor of olaparib monotherapy in the adjuvant setting after completion of chemotherapy. That study is going to change the way we treat patients with germline BRCA1/2 mutations.

Has alpelisib (Piqray) become the standard treatment for patients with HR-positive disease and PIK3CA mutations based on data from the phase 3 SOLAR-1 trial (NCT02437318)?

Alpelisib is a PI3K inhibitor. SOLAR-1 tested alpelisib plus fulvestrant in patients who had previously received an AI. The subset of patients with PIK3CA mutations derived the [most] benefit. Patients without those mutations didn’t seem to have benefit with alpelisib. The progression-free survival [PFS] was about 6 months [in patients with the mutation who received placebo plus fulvestrant] vs 11 months in [patients who received the combination]. At the 2020 ESMO Congress, the investigators reported the OS update with about an 8-month difference in favor of alpelisib, which did not reach statistical significance.

We also saw a subset analysis of circulating tumor DNA (ctDNA)-positive patients at baseline, and there was a bigger effect in them; there, we saw about a 9-month difference in favor of alpelisib. There may be a way to tailor this [drug] a little bit more directly because it’s a hard drug [to tolerate]. Hyperglycemia, gastrointestinal adverse effects [AEs], and rash can be troublesome for these patients.

There are a couple of drugs that we have available that target this pathway. There’s alpelisib, which inhibits a subunit component of PI3K, and there’s everolimus [Afinitor], which has been studied in randomized trials. Also, [everolimus is] an mTOR inhibitor, so it inhibits mTORC1. The key trials, SOLAR-1 and BOLERO-2 [(NCT00863655) enrolled] relatively similar populations and demonstrated very similar effects on PFS. The absolute difference in OS is a little bit different.

An unanswered question here is: Who’s better for alpelisib and who’s better for a more downstream drug like everolimus? The AEs are different, and that hasn’t yet been sorted out.

How do you currently approach that decision in practice?

The everolimus data do not suggest an interaction with mutations in PIK3CA, so [the drug] seemed to work regardless of those features, unlike alpelisib, which is a much more specific drug. In my practice, I use alpelisib in patients I know have mutations, and I will use everolimus in a more global population, or in those who don’t tolerate alpelisib. Hyperglycemia can be quite hard to manage in these patients. Now, that can happen with both drugs, but it feels like it’s a little harder with alpelisib.

In TNBC, how has our understanding of the activity of sacituzumab govitecan-hziy (Trodelvy) grown since its approval in 2020?

The main ADC that we all are using and has had wonderful data for is sacituzumab govitecan. It has a Trop-2–directed antibody and a topoisomerase payload. It was tested in a seminal trial called ASCENT [NCT02574455] in pretreated TNBC compared with physician’s choice chemotherapy. The parent trial has been published and was the underpinning for the FDA approval of this drug because it was strongly in favor of the ADC with a PFS that was essentially tripled from about 2 months to 6 months and an OS advantage from about 7 months to about 12 months in heavily pretreated patients. These patients on average had received 4 prior cytotoxic regimens, so this is a very effective agent.

What was updated at the 2021 ASCO Annual Meeting was the subset of patients who had a very early relapse. There were about 65 patients in the parent trial who had early relapse within a year and had received only 1 regimen for metastatic disease, so they’re less pretreated, but also have features of essentially chemotherapy resistance because they relapse quickly. The drug worked similarly in these patients as it did in the more heavily pretreated but less chemotherapy-resistant population of the parent trial, with an OS advantage somewhere around 6 months and a PFS difference from 1.5 months to close to 6 months. The effect appeared to be very similar to the parent trial. In TNBC, these patients with early relapses don’t appear to be resistant to a cytotoxic delivered in this mechanism, with the ADC sort of acting as the Trojan horse to deliver the cytotoxic into the Trop-2–expressing cancer cells.

Are there any ongoing efforts that are being made to further improve on the efficacy that has been seen with sacituzumab govitecan?

A lot of ideas are swirling around, and they largely relate to how sacituzumab govitecan does in an earlier setting. How does it do against cytotoxics in a more conventional early-line setting and in a less heavily pretreated setting? The hint from ASCO is that it’s going to outperform [its comparator] in an earlier-line setting. Moving [the agent] into the first-line [metastatic] setting, looking [into its use in] the residual disease setting, and looking into [its use in] early breast cancer where curative-intent therapy is the goal [is our aim]. Those are all going to be questions that will be important for [us to answer].

Where might datopotamab deruxtecan (DXd) fit into the mix?

DXd is another Trop-2–directed ADC with a cleavable linker and a topoisomerase 1 payload. It’s shown activity in lung cancer and is now being studied in a bunch of tumors, including breast cancer. At the 2021 ESMO Breast Congress, the triple-negative subset of the pan-tumor trial was presented on the 21 patients that were evaluable. These were heavily pretreated patients who had received about 4 prior lines of therapy and experienced a 43% response rate, so it really looks like [we’re seeing] nice evidence of activity.

The problem with the DXd-type ADCs is that the benefit and the risk are associated with the nature of the linkage that’s a cleavable linker, so it has a bystander effect. That’s why fam-trastuzumab deruxtecan-nxki [Enhertu] has a lot of effectiveness but it may also be why it has a lot of interstitial lung disease [ILD]. Now, in these 21 patients, the investigators didn’t see any ILD, but these are the kinds of things that we need to watch for.

One of the presentations focused on the practical applications of genomic assays in early-stage, ER-positive breast cancer. What should be taken away from that talk?

Dr. Emily Ray gave that presentation, and she focused on the assay that we use most commonly in the United States: OncotypeDX. She commented, appropriately, that at lower recurrence scores the use of chemotherapy can be omitted. There is very excellent data in the node-negative setting for postmenopausal patients with a recurrence score of up to 25 to omit chemotherapy. Somewhere between 16 and 20 is when the curves tended to cross in the premenopausal setting in what is admittedly a subset analysis that isn’t terribly robust but is consistent with our biologic assumptions.

She also pointed out the same kind of interaction with menopausal status in node-positive tumors, meaning that you must tailor your use of these genomic assays, according to what the data show you. In postmenopausal patients, there was no evidence of a benefit of chemotherapy. The investigators tested up to 25 in the RxPONDER trial [NCT01272037], but in premenopausal patients, [chemotherapy] didn’t help anyone, and it didn’t seem to matter by recurrence score up until 25, which was the limit of what the trial tested.

In premenopausal patients, the chemotherapy seemed to have a benefit independent of the recurrence score, so I’m not sure there’s a particular reason to order a test like this in premenopausal patients with node-positive breast cancer, where the chemotherapy seems to have a role to play regardless of [recurrence score].

There are differences amongst these assays in terms of how much is driven by proliferation and how much is driven by ER signaling. Whether a different assay would perform better is a reasonable question and one that wasn’t addressed by the RxPONDER trial.

What other trials have read out that have evaluated the necessity of chemotherapy?

There was a Chinese trial, SYSUCC-002 [NCT01950182], which asked a very practical question in patients who have both HR and HER2 positivity on their tumor. This study compared endocrine therapy plus trastuzumab vs chemotherapy plus trastuzumab in the first-line setting.

The investigators were looking for noninferiority of the endocrine-based approach compared with chemotherapy, and the hazard ratio was in favor of the endocrine therapy in this group of about 400 patients with [luminal B2 breast cancer]. The one group that seemed to do better with chemotherapy were those that had a very short DFI of less than 2 months, at which point, the chemotherapy looked like it outperformed [the endocrine therapy] in these subset analyses. In general, patients treated with endocrine therapy up front, seem to do well with anti-HER2 therapy. In this study, it was just trastuzumab, but I can’t think of a reason why it wouldn’t be similar with trastuzumab and pertuzumab.

In HER2-positive breast cancer, what was the significance of the findings from the FeDeriCa trial (NCT03493854)?

This is just a huge boon for our patients. The FeDeriCa trial was a was a seminal trial testing the subcutaneous formulation of trastuzumab [Herceptin] and pertuzumab [Perjeta; HP], which is used conventionally and given intravenously [IV]. FeDeriCa was a very straightforward comparison [of both formulations] in a neoadjuvant setting. It [evaluated] an anthracycline/taxane, conventional AC-THP–type regimen, in which the HP part in one arm was given IV and the other arm was given subcutaneously. Whatever patients were assigned to, they continued that into the adjuvant phase to complete their year of anti-HER2 therapy.

The trial was a noninferiority trial because it tested a subcutaneous vs IV formulation. The primary end point was the serum trough concentration of pertuzumab, which from a clinician standpoint is less important, but from a pharmacologist standpoint, is important. The trial essentially met its end points for this trough [concentration]. From a clinical standpoint, the investigators also looked at pathologic complete response rates, and they were essentially identical, meaning that the subcutaneous version of these drugs is perfectly appropriate and reasonable to use.

There’s another study called PHranceSCa trial (NCT03674112) that’s been reported as a randomized trial in the adjuvant setting. The study really had to do with patient preference, and patients by far preferred the subcutaneous formulation, which is easy to administer and there’s less time [spent in the infusion chair], so it’s a win-win.