HR+ BC Insights: SABCS 2019
New Developments in HR-Positive/HER2-Negative Breast Cancer: Insights From SABCS 2019

Released: January 31, 2020

Expiration: January 29, 2021

Activity

Progress
1
Course Completed

The 2019 San Antonio Breast Cancer Symposium (SABCS) was a productive and exciting meeting where we saw several studies report promising new treatment strategies poised to improve outcomes for patients with breast cancer. In this commentary, I focus on advances in hormone receptor (HR)–positive/HER2-negative disease, including 10-year follow-up data from the phase III NSABP B‑42 trial evaluating extended adjuvant letrozole in postmenopausal women with HR-positive, early-stage breast cancer (EBC) after previous adjuvant aromatase inhibitor (AI) therapy; the phase III PEARL trial comparing the CDK4/6 inhibitor palbociclib plus endocrine therapy vs capecitabine head to head in postmenopausal patients with HR-positive/HER2-negative metastatic breast cancer (MBC) and disease recurrence or progression on AI therapy; and, the plasmaMATCH proof-of-concept trial designed to assess the clinical utility of analyzing circulating tumor DNA (ctDNA) to inform therapy decisions in patients with MBC. In companion commentaries, my colleagues, Sara Hurvitz, MD, FACP, and Sara Tolaney, MD, MPH, discuss exciting new practice-changing data in HER2-positive breast cancer and new developments in chemotherapy-based treatment for HER2-negative and triple-negative breast cancer from SABCS 2019.

Phase III NSABP B‑42 10-Year Follow-up: Extended Adjuvant Letrozole After Previous Adjuvant AI Therapy in Postmenopausal Women With HR-Positive EBC
Among the most highly anticipated reports at SABCS 2019 was the 10‑year follow-up results from the randomized, double-blind phase III NSABP B‑42 trial. This important clinical trial assessed extended adjuvant endocrine therapy in postmenopausal women with HR‑positive/HER2‑negative EBC. Patients who were disease-free after 5 years of treatment with an AI or 2-3 years of treatment with tamoxifen followed by AI therapy (totaling at least 5 years of endocrine therapy) were randomized to receive 5 additional years of letrozole or placebo (N = 3966). The primary endpoint was disease-free survival (DFS); secondary endpoints included OS, breast cancer–free interval, and distant recurrence.

The primary analysis with a median follow-up of 6.9 years showed a hazard ratio of 0.85 (95% CI: 0.73-0.999) for DFS with letrozole vs placebo but did not quite meet statistical significance (P = .048). However, extended letrozole therapy was associated with a statistically significant improvement in breast cancer–free interval, with a hazard ratio of 0.71 (95% CI: 0.56-0.89; P = .0027), corresponding to a 29% reduction in the risk of recurrence.

Now at the 10‑year mark with a median follow-up of 9.3 years, the primary endpoint of DFS was significantly improved in patients who received extended letrozole vs those who received placebo, with a hazard ratio similar to the primary analysis at 0.84 (95% CI: 0.74-0.96) but now with a P value of .011. Patients in the letrozole arm also had reduced rates of distant recurrence (4.2% vs 5.7%) and second primary breast cancer (2.7% vs 4.1%). Although there was no difference in OS with extended letrozole vs placebo, there were significant improvements in breast cancer–free interval (hazard ratio: 0.74; 95% CI: 0.61-0.91; P = .003) and, of most importance, distant recurrence (hazard ratio: 0.71; 95% CI: 0.55-0.93; P = .04), consistent with previous findings from the primary analysis at 7 years. Of interest, the curves for breast cancer–free interval and distant recurrence are not yet parallel at 10 years and appear to be continuing to separate.

Another interesting finding came from the subgroup analysis according to bone mineral density. Patients with a T score ≤ -2 in hip or spine—that is, patients who were substantially osteopenic or osteoporotic—demonstrated an impressive DFS benefit with extended letrozole as compared with patients with higher T scores (P = .01). The mechanism driving this difference is unclear. However, this finding that patients with low bone mineral density may benefit more from extended adjuvant letrozole may not be pertinent to today’s practice where the standard of care is to recommend bisphosphonates to support bone density and to improve outcomes from breast cancer in postmenopausal patients receiving endocrine therapy.

In addition, there was a numerical improvement in DFS with extended letrozole among patients with node-positive disease and those who previously received tamoxifen. Of interest, recent guidance from ASCO recommends that patients with node‑positive disease should receive 10 years of adjuvant endocrine therapy, which is in agreement with the numerical improvement in DFS seen in this subgroup of patients in the NSABP B‑42 trial.

The take-home conclusion from NSABP B-42 is that extended adjuvant endocrine therapy is beneficial in select higher‑risk patients with HR‑positive/HER2‑negative EBC. Based on these data, and in accordance with current guidelines, I would use extended adjuvant endocrine therapy in patients with higher-risk disease, including node-positive, stage III disease, or patients with large or biologically aggressive tumors in the setting of node-negative disease. In practice, I already have been counseling patients with node‑positive disease that 10 years of therapy is ideal based on earlier clinical trials, including the MA.17R trial. For patients with higher‑risk, node-negative disease, I am generally also continuing adjuvant endocrine therapy beyond 5 years. Because most patients do not experience significant adverse events from the medication after a few years on therapy, they tend to be amenable to remaining on therapy for longer durations when recommended.

Phase III PEARL: Palbociclib Plus Endocrine Therapy vs Capecitabine in Postmenopausal Women With HR-Positive/HER2-Negative MBC and Previous AI Therapy
The randomized, international phase III PEARL trial was an important head-to-head comparison of a CDK4/6 inhibitor plus endocrine therapy vs chemotherapy in postmenopausal patients with HR-positive/HER2-negative MBC (N = 601). Participants in this trial were pretreated, with all having received an AI (whether for EBC or MBC) and most having previously received chemotherapy and/or endocrine therapy for MBC. Of note, the majority of patients had visceral disease (eg, liver or lung metastases), which is a setting where physicians continue to wonder whether it is safe to use a CDK4/6 inhibitor plus endocrine therapy or if is better to go with capecitabine, a chemotherapy agent with substantial activity in the liver.

The PEARL trial had a modified study design where 601 participants were enrolled in 1 of 2 cohorts. Those in cohort 1 were randomized to receive palbociclib plus exemestane vs capecitabine (n = 296); enrollment for this cohort occurred between March 2014 and September 2016. Subsequently, a second cohort was added based on new, mostly preclinical data identifying a role for ESR1 mutations in mediating AI resistance and identifying fulvestrant as potentially more active against tumors with ESR1 mutations. Cohort 2 participants, who enrolled between May 2016 and July 2018, were randomized to receive palbociclib plus fulvestrant vs capecitabine (n = 305). Only patients with no previous capecitabine, and no previous exemestane or fulvestrant (depending on the cohort), for MBC were permitted. The coprimary endpoints were PFS among all patients with wild-type ESR1 tumors in both cohorts and PFS in cohort 2 regardless of ESR1 mutation status. Secondary endpoints included PFS regardless of ESR1 mutation status, OS, ORR, and safety.

Analysis of the coprimary endpoints across cohorts revealed no statistically significant difference in PFS with palbociclib plus endocrine therapy vs capecitabine. Among patients with wild-type ESR1 tumors in both cohorts, the median PFS with palbociclib plus endocrine therapy was 8.0 months vs 10.6 months with capecitabine (hazard ratio: 1.08; 95% CI: 0.85-1.36; P = .526). Among all patients in cohort 2 (including those with wild-type ESR1 and mutant ESR1 tumors), patients who received palbociclib plus fulvestrant had a median PFS of 7.5 months vs 10.0 months with capecitabine (hazard ratio: 1.09; 95% CI: 0.83-1.44; P = .537). Analysis of the secondary PFS endpoint in all patients regardless of ESR1 mutation status also found no difference between treatment arms (hazard ratio: 1.09; 95% CI: 0.90-1.31; P = .380).

It is interesting to note that the PFS benefit with either regimen was the same regardless of ESR1 mutation status. Some clinicians may think that chemotherapy is the best option for patients with an ESR1 mutation, out of concern that endocrine therapy will not be as effective. However, these results show that a CDK4/6 inhibitor combined with endocrine therapy is just as effective as chemotherapy in the context of ESR1 mutations. This was also the case for visceral disease. In patients with visceral disease in cohort 2, the hazard ratio for PFS with palbociclib plus fulvestrant vs capecitabine was 1.04 (95% CI: 0.75-1.45). In patients with wild-type ESR1 tumors and visceral disease in both cohorts, the hazard ratio for PFS with palbociclib plus endocrine therapy vs capecitabine was 1.00 (95% CI: 0.75-1.34). For any physicians still skeptical about using a CDK4/6 inhibitor plus endocrine therapy in patients with visceral disease, these data should assure them that even these patients will achieve good outcomes on a CDK4/6 inhibitor–based regimen. Certainly, in the frontline setting, we know that CDK4/6 inhibitors will provide longer PFS than chemotherapy, even in patients with visceral disease.

Moreover, the toxicity is lower with a CDK4/6 inhibitor plus endocrine therapy compared with chemotherapy. For example, in PEARL, there were fewer serious adverse events and fewer adverse events leading to discontinuation in the palbociclib plus endocrine therapy arms than with capecitabine.

Based on these findings, we can comfortably conclude that there is no compelling reason to administer chemotherapy in patients with HR-positive/HER2-negative MBC previously treated with an AI, particularly for patients with good end‑organ function and good performance status such as those who enrolled on the PEARL trial. I think this head-to-head comparison provides us with very helpful, practical data for our practice.

plasmaMATCH (CRUK/15/010): Evaluation of ctDNA Testing to Direct Targeted Therapies in MBC
The plasmaMATCH trial was a very important and much needed proof-of-concept trial designed to assess the clinical utility of analyzing ctDNA to identify targetable mutations and inform therapy decisions in patients with MBC. This trial enrolled close to 1000 patients with locally recurrent or metastatic breast cancer and measurable disease who had relapsed within 12 months of adjuvant chemotherapy or had progressive disease following previous treatment for advanced disease. Participants were permitted to have received up to 2 prior chemotherapy regimens and had to have an actionable mutation as detected by ctDNA testing.

The plasmaMATCH trial assigned participants to 1 of 5 cohorts based on tumor mutation (Figure). Cohort A enrolled patients with an ESR1 mutation to receive high‑dose fulvestrant. Cohort B enrolled patients with a HER2 mutation to receive the pan-HER inhibitor neratinib with the addition of standard-dose fulvestrant if their breast cancer was estrogen receptor positive. Cohort C enrolled patients with an activating AKT1 mutation whose disease was also estrogen receptor positive; they received the oral AKT inhibitor capivasertib plus standard-dose fulvestrant. Patients with estrogen receptor–negative breast cancer who had an AKT or PTEN mutation were assigned to cohort D, where they received single-agent capivasertib. Finally, patients with triple‑negative breast cancer with no actionable mutations were assigned to cohort E, where they received an investigational regimen consisting of olaparib plus the ATR kinase inhibitor, AZD6738.

Figure. plasmaMATCH Study Design

###PLACEHOLDER###

 

The primary objective of the study was to assess the response to targeted therapies that were matched to mutations identified by ctDNA testing, with the primary endpoint being ORR by RECIST v1.1. Secondary objectives included determining the frequency of actionable mutations and assessing the accuracy of ctDNA testing, with secondary endpoints including duration of response, clinical benefit ratio, and PFS.

In plasmaMATCH, ctDNA testing for mutations was performed on plasma samples that had been collected upon progression on patients’ most recent treatment. ctDNA testing was performed in 2 different ways, next-generation sequencing and digital droplet PCR, an older technology, in 2 different laboratories. ESR1 mutations were the most frequent mutations identified, occurring in 27.7% of patients. The incidence of HER2 and AKT1 mutations was 2.7% and 4.2%, respectively. Of note, the investigators did not test for mutations in PIK3CA because the PI3K inhibitor alpelisib is already approved in combination with fulvestrant as a standard-of-care regimen for patients with these mutations. There was good concordance in mutations identified by next-generation sequencing and digital droplet PCR. Furthermore, good concordance was found when the investigators compared mutations identified via plasma ctDNA testing with matched tumor samples, including primary or metastatic tumor tissue. This was reassuring to see and suggests that ctDNA testing is a reliable method for identifying tumor mutations.

To summarize the main results from each cohort, let’s begin with cohort A. Among the 74 patients with ESR1 mutations who received high-dose fulvestrant, the confirmed response rate was 8.1% and the median PFS was 2.2 months—not remarkable results but some patients benefitted. 

In the 20 patients with HER2 mutations in cohort B, neratinib with or without fulvestrant achieved a sizeable level of activity, with a confirmed response rate of 25% and median PFS of 5.4 months. For the 17 patients whose disease was also HR positive, which was the majority, the response rate was 23.5%. Considering these response data along with updated results from the phase II SUMMIT trial—which reported a confirmed response rate of 53% with neratinib plus trastuzumab and fulvestrant among patients with HR-positive and HER2-mutated MBC—I consider neratinib to be an effective option for patients with activating HER2 mutations, particularly when also inhibiting the estrogen receptor with fulvestrant. Based on these data, I hope to see FDA approval for neratinib in this setting so we can use this routinely in our practice for these patients.

Among the 18 patients in cohort C, who had an AKT1 mutation and received capivasertib plus fulvestrant, the confirmed response rate was 22% and the median PFS was 10.2 months. This was an impressive level of activity consistent with previous findings of capivasertib plus fulvestrant efficacy in patients with and without activating AKT pathway mutations, including in the phase II FAKTION trial.

Finally, in cohort D, the AKT activation basket arm, single-agent capivasertib achieved a confirmed response rate of 10.5% and a median PFS of 3.4 months in the 19 patients with estrogen receptor–negative advanced breast cancer. In patients who had an AKT mutation, 2 of 6 (33%) patients had a response whereas no patients with a loss of function PTEN mutation did. Correspondingly, capivasertib is being evaluated in combination with paclitaxel as first-line treatment for patients with locally advanced or metastatic triple-negative breast cancer in the phase III Capltello209 trial.

I was pleased to see the results of the plasmaMATCH trial because they demonstrate that we can use plasma to test for mutations that can be targeted by modern therapies, whether available in practice now or within a clinical trial. Being able to test plasma samples instead of having patients undergo tumor biopsies for mutation analysis will be very useful when testing for actionable mutations becomes standard practice in breast cancer care, which I think is in our not-too-distant future considering that approximately 50% of our patients with breast cancer will have an actionable mutation, whether in ESR1, PIK3CA, AKT, or HER2. In fact, physicians are now becoming accustomed to screening patients for PIK3CA mutations in the wake of the approval of alpelisib, which is indicated for use in patients whose cancer has a PIK3CA mutation, is also HR positive/HER2 negative, and whose disease has progressed on AI therapy. The current test for PIK3CA mutations is based on tissue but there is a ctDNA-based test forthcoming.

More SABCS 2019 Conference Coverage on the CCO Web Site!
These are just a few of the interesting and important abstracts selected by the expert faculty from SABCS 2019. Downloadable slideset summaries of these studies and more are available on our Web site, including additional CME-certified commentaries discussing recent findings in HER2-positive breast cancer, new developments in chemotherapy-based treatment for HER2-negative and triple-negative breast cancer, and encouraging data on neoadjuvant immune checkpoint blockade in early breast cancer.

Poll

1.
Will the results from the phase III PEARL trial change how you use chemotherapy for your patients with HR-positive/HER2-negative MBC and visceral disease?
Submit