Endocrine-Sensitive Disease Rate in Postmenopausal Patients With Estrogen Receptor–Rich/ERBB2-Negative Breast Cancer Receiving Neoadjuvant Anastrozole, Fulvestrant, or Their Combination

This randomized clinical trial assesses whether neoadjuvant fulvestrant or anastrozole plus fulvestrant vs anastrozole alone increases endocrine-sensitive disease rate (ESDR) in postmenopausal women with estrogen receptor (ER)–rich/ERBB2-negative breast cancer.

2][3][4][5] Selective ER degraders (SERDs) have the potential to further improve outcomes by inhibiting both estrogen-dependent and estrogenindependent ER signaling. 6The SERD fulvestrant improved progression-free survival compared to anastrozole as firstline therapy for postmenopausal women with advanced hormone receptor (HR)-positive BC in the FALCON trial. 7The combination of anastrozole plus fulvestrant (A+F) was also investigated in the metastatic setting based on preclinical evidence that fulvestrant is more effective in a low-estrogen environment, 8 and that AI plus fulvestrant was more effective than monotherapy. 9,10The SWOG S0226 trial demonstrated that A+F improved survival compared to anastrozole as first-line therapy for advanced HR-positive BC.Benefits were particularly notable in the endocrine therapy (ET)naive population. 11,12In contrast, the FACT trial found no significant improvement in time to progression with the addition of fulvestrant to anastrozole in the first-line metastatic setting, but few patients were ET naive. 13Both trials evaluated fulvestrant at 250 mg, raising the question of whether anastrozole would add to the efficacy of fulvestrant at the currently approved 500-mg monthly dose. 14,15eoadjuvant ET (NET) trials provide opportunities to assess individual endocrine sensitivities in treatment-naive early-stage ER-positive/ERBB2-negative BC. 16 ET resistance can be defined by tumor Ki67 greater than 10% following 2 to 4 weeks of NET, which is a prospectively validated biomarker for increased recurrence risk in postmenopausal individuals. 17In addition, a higher degree of Ki67 suppression in randomized NET trials predicted improved adjuvant efficacy of AI compared to tamoxifen. 16The finding that pathological stage, residual tumor Ki67, and ER levels following 4 to 6 months of neoadjuvant AI or tamoxifen treatment were all independent prognostic factors led to the development of the preoperative endocrine prognostic index (PEPI) to determine the risk of relapse. 18A PEPI 0, defined as ypT1-2 N0, residual tumor ER positive, and Ki67 of 2.7% or less, was associated with a 5-year relapse risk less than 5%. 18,19PEPI therefore offers a novel end point for NET trials, as pathologic complete response (pCR) is rare.In designing the ALTERNATE (Alliance A011106) trial reported here, we used modified PEPI (mPEPI) to exclude the ER component because fulvestrant reduces ER levels.The mPEPI remained prognostic in previous NET trials because loss of ER to below an Allred score of 3 was rarely the exclusive cause of PEPI greater than 0. 18,20 Herein is the primary analysis of the neoadjuvant phase of the ALTERNATE trial designed to determine whether fulvestrant alone or A+F increases the rate of endocrinesensitive disease (ESD), defined as pCR or mPEPI 0 (ypT1-2N0/N1mic/Ki67 ≤2.7%).Secondary end points included Ki67 suppression after 4 weeks of NET and the pCR rate after switching to neoadjuvant chemotherapy (NCT) due to a week 4 or week 12 Ki67 greater than 10%.As ER-positive BC is molecularly heterogeneous, ALTERNATE correlative studies included preplanned genomic and transcriptomic analyses to gain insights into ET resistance mechanisms.Since ACOSOG Z1031 demonstrated that the rate of PEPI 0 and Ki67 suppression from neoadjuvant AI differed by PAM50 subtype, 21 the association between RNA sequencing-based PAM50 subtype and week 4 Ki67 suppression was explored.

Study Design
This multi-institutional open-label phase 3 randomized clinical trial enrolled postmenopausal women with treatmentnaive palpable clinical T2-T4c, any N, M0, ER-rich (Allred score 6-8 or >66%), ERBB2-negative BC.Full eligibility criteria can be found in protocol section 3.0 (Supplement 1).Each participant was asked to select her race and ethnicity from among the National Cancer Institute-defined categories.Participants were free to not report this information or to specify something other than the predefined categories.Patients were randomly assigned (1:1:1) to the anastrozole, fulvestrant, or A+F arm, using the Pocock-Simon dynamic allocation procedure, balancing the marginal distributions of the stratification factors between arms.The stratification factors were clinical tumor stage (T2 vs T3 vs T4a-c), lymph node status (positive vs negative), and Eastern Cooperative Oncology Group performance status (0-1 vs 2).Treatments were anastrozole, 1 mg by mouth daily; fulvestrant, 500 mg intramuscularly on days 1 and 15 of cycle 1, then day 1 of subsequent cycles; or A+F (same doses) for 6 cycles (cycle length: 28 days) followed by surgery (eFigure 1 in Supplement 2).Research BC biopsies were required pretreatment, at week 4, and at the time of surgery, and optionally at week 12 for central Ki67 assessment.Patients with week 4 or week 12 Ki67 greater than 10% were to switch to NCT or surgery.NCT regimens were paclitaxel for 12 weeks or per standard of care.Patients with week 4 or week 12 Ki67 10% or less or insufficient tumor cells to determine week 4 or week 12 Ki67 continued their assigned NET.Bidimensional breast lesion measurements were performed on day 1 of each cycle prior to surgery.Progression on examination was confirmed by mammography or ultrasonography.Breast surgery was per standard of care.

Key Points
Question Is fulvestrant, alone or in combination with the aromatase inhibitor anastrozole, superior to anastrozole alone as neoadjuvant endocrine therapy (NET) in postmenopausal women with estrogen receptor (ER)-rich/ERBB2-negative breast cancer, in terms of endocrine-sensitive disease rate (ESDR)?
Findings In this phase 3 randomized clinical trial of 1362 patients, the ESDR following 6 months of neoadjuvant fulvestrant or anastrozole plus fulvestrant was 22.8% and 20.5%, respectively; neither was significantly higher than the 18.7% ESDR with anastrozole alone.
Meaning Aromatase inhibition remains the standard-of-care NET for postmenopausal ER-rich/ERBB2-negative breast cancer.
Sentinel lymph node with or without axillary lymph node dissection was required to ensure ability to determine PEPI score and residual cancer burden (RCB). 22Central Ki67 and ER assessment and RNA extraction and sequencing for PAM50 subtype are described in Supplement 1.The trial was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines.The protocol was approved by the National Cancer Institute Adult-Late Phase central institutional review board (IRB) and local IRBs as appropriate.Each participant signed an IRB-approved, protocol-specific informed consent document in accordance with federal and institutional guidelines.This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

Statistical Analysis
All eligible patients who began protocol treatment were evaluable and included in the analysis cohort (Figure).The primary outcome was ESDR defined as the proportion of patients with pCR (no invasive disease in breast or lymph nodes) or mPEPI 0. Patients with week 4 or week 12 Ki67 greater than 10%, confirmed progressive disease (PD), mPEPI greater than 0, or insufficient data to determine mPEPI or who discontinued protocol treatment for any reason, without completing surgery, were considered as not having ESD.
The trial was designed to ascertain whether fulvestrant alone or A+F increases the ESDR by 10% or greater over anastrozole alone.The sample size was determined under the assumption that ESDR with anastrozole alone would be similar to that in the ACOSOG Z1031B trial, namely 30%. 21With a sample size of 425 per arm, a 1-sided α of 0.025 χ 2 test of 2 independent proportions would have 84% power to detect an increase of 10% or greater in ESDR for a given fulvestrantcontaining arm relative to that in the anastrozole arm.A Wald test, which is asymptotically equivalent to this χ 2 test, was used for assessing the primary aim.No adjustments for randomization stratification factors were made.
Wilcoxon rank sum tests were used to assess whether the percentage change in Ki67 after 4 weeks of NET differed between treatments and by PAM50 subtype.Conditional logistic regression modeling with the Score method of model selection was used with pretreatment Ki67 as the stratification factor to assess whether the likelihood of week 4 Ki67 greater than 10% differed by arm and then with treatment arm as the stratification factor to assess whether age, body mass index (calculated as weight in kilograms divided by height in meters squared), Eastern Cooperative Oncology Group performance status, cT stage, cN stage, grade, or pretreatment Ki67 were associated with likelihood of not achieving ESD.Testing was performed such that a 2-sided P value less than .05was considered significant.The preplanned end point for the cohort of women who switched to NCT due to week 4 or week 12 Ki67 greater than 10% was pCR/RCB class I (RCB-I) rate, defined as the proportion with pCR or RCB-I.Analyses were performed using SAS, version 9.4M6 (SAS Institute).All analyses were based on data frozen on March 2, 2023.Information regarding data and safety monitoring and data quality is provided in Supplement 1.

Patient Characteristics
From February 2014 to November 2018, 1362 female patients (mean [SD] age, 65.0 [8.2] years) were registered to the neoadjuvant phase.Subsequently, 38 were found to be ineligible, and 26 withdrew consent prior to starting protocol treatment.The remaining 1298 (anastrozole: n = 434; fulvestrant: n = 430; A+F: n = 434) women composed the analysis cohort (Figure).Baseline characteristics were similar across the arms (Table 1).Most patients were White, more than 40% had node-positive disease, and approximately 30% had locally advanced T3-4c disease.At baseline, 10% to 15% of patients had pretreatment Ki67 10% or less, and few (approximately 2% across all arms) had pretreatment Ki67 2.7% or less (Table 1). .Each NET regimen was well tolerated, with similar adverse event profiles (eTable 1 in Supplement 2).The grade 3 to 4 adverse event rate ranged from 4% to 6% across the 3 NET arms.

Discussion
In this phase 3 randomized clinical trial, the 6-month ESDR following anastrozole, fulvestrant, or A+F was 19% to 23% in Neither fulvestrant nor A+F significantly increased the ESDR over anastrozole.There was a low rate of pCR (0.8%) and axillary lymph node clearance (10%) across the arms, which was consistent with previous reports. 19,21,23This study demonstrated that nodal positivity, locally advanced tumor size, high grade, or pretreatment Ki67 greater than 20% significantly increased the likelihood of not achieving ESD.A higher proportion of patients had node-positive and locally advanced disease in ALTERNATE than in the Z1031B trial, which may have contributed to the lower than expected ESDR. 19However, PD on NET was rare in this trial, approximately 1% overall.This may be a consequence of the early triaging of patients with on-treatment Ki67 greater than 10% off NET. 24 In contrast to limited pathologic downstaging, complete cell cycle arrest (Ki67 ≤ 2.7%) at surgery was observed in 43.1% of the 1272 patients with baseline Ki67 greater than 2.7%, indicating the potent antiproliferative effects of NET.It is noted that the likelihood of week 4 or week 12 Ki67 greater than 10% was lower on A+F than anastrozole, leading to fewer patients receiving A+F going off NET at week 4 than those receiving anastrozole.However, this did not translate to an improved ESDR or significantly greater Ki67 suppression at week 4 between A+F and anastrozole because of the limited effect on the other, histopathological, PEPI components.
As a SERD, the antitumor activity of fulvestrant has been shown to be dependent on its ability to degrade ER. 25 The observation that ER remained high (Allred score 6-8) in more than 70% of patients after 6 months of fulvestrant or A+F in this trial indicates that fulvestrant is a relatively ineffective ER degrader.While dose-dependent ER degradation was observed for fulvestrant in the NEWEST trial, 26 the ability to dose escalate fulvestrant is limited by pharmacokinetic properties.Our data therefore support the development of more effective SERDs or other novel ER-targeting mechanisms. 27The recent approval of elacestrant for ESR1-mutated metastatic BC represents an example; however, ESR1 mutations are rarely present at diagnosis. 28he 4.8% pCR rate after switching to NCT due to week 4 or week 12 Ki67 greater than 10% is consistent with observations in the Z1031B trial. 19The much larger sample size of ALTERNATE provides further support for the relative ineffectiveness of chemotherapy for ET-resistant tumors.The RCB-I rate was also low.Since post-NCT RCB index is prognostic in patients with BC, including those with HR-positive/ERBB2negative disease, 22,29 our data underscore the need for more effective systemic treatments for these patients.
PAM50 subtype, which is the basis for the commercial assay Prosigna, is prognostic in ER-positive BC, 30,31 and emerging evidence indicates an association with ET sensitivity. 21In ALTERNATE, A+F led to a significantly greater Ki67 suppression and decreased likelihood of maintaining Ki67 greater than 10% at week 4 in LumB vs LumA tumors.This finding is hypothesis generating, but suggests that despite relative AI resistance, LumB tumors, which comprise 40% of ER-rich BCs, are frequently ER dependent.Thus, improvements in ET may particularly benefit this tumor subset.Notably, nonluminal BC was found in 7.3% of patients in the PAM50 cohort despite high ER levels required for eligibility.Their markedly reduced sensitivity to NET is consistent with earlier findings in the Z1031 trial, 19 thus calling into question the appropriateness of NET for nonluminal ER-positive BCs.

Strengths and Limitations
To our knowledge, ALTERNATE is the first reported clinical trial comparing fulvestrant or A+F with anastrozole in postmenopausal women with early-stage ER-rich/ERBB2negative BC.Strengths include the randomized phase 3 trial design, central Ki67 analysis, early triaging to chemotherapy/ surgery if week 4 or week 12 Ki67 greater than 10%, and more than 750 tumors analyzed by PAM50.There are several limitations.First, the importance of achieving ESD (mPEPI 0) depends on whether it is associated with a low risk of disease recurrence.This question is being examined in the adjuvant portion of the ALTERNATE trial.However, further follow-up is required to address the relapse-free survival end point.
Second, radiographic response rates were not analyzed since pretreatment and posttreatment imaging was completed in less than 60% of patients.Lastly, baseline clinical characteristics of the PAM50 cohort, representing 58% of the trial population, were similar between the treatment arms, but had disease more locally advanced, higher in grade, and higher in pretreatment Ki67 than those without PAM50 determination.

Conclusions
In this randomized clinical trial, the neoadjuvant phase of the ALTERNATE trial did not demonstrate superiority for fulvestrant or A+F over anastrozole alone in improving ESDR.Pathologic eradication of disease was uncommon across the arms, largely limiting mPEPI 0 status following NET to those with initial node-negative disease at presentation.The limited ER degradation following fulvestrant or A+F supports the ongoing development of more effective SERDs.ALTERNATE also confirmed the poor pathologic response to NCT among ER-rich/ERBB2-negative BC resistant to NET.The higher antiproliferative effect of combining AI with a SERD in the LumB population is provocative and warrants further study, particularly in light of the development of oral SERDs. 26 Of the 1298 eligible patients who began NET, 933 (71.9%) (anastrozole: n = 289; fulvestrant: n = 306; A+F: n = 338) had week 4 or week 12 Ki67 10% or less or insufficient tumor cells to ascertain week 4 or week 12 Ki67 and completed 6 cycles of NET and surgery (Figure).A total of 302 patients (23.2%) discontinued NET by study design due to week 4 or week 12 Ki67 greater than 10% (anastrozole: n = 109; fulvestrant: n = 104; A+F: n = 68) or PD by imaging (anastrozole: n = 7; fulvestrant: n = 6; A+F: n = 8).Another 63 (4.9%) discontinued during NET or prior to surgery due to patient choice (anastrozole: n = 16; fulvestrant: n = 9; A+F: n = 10), adverse event (anastrozole: n = 4; fulvestrant: n = 1; A+F: n = 1), or other reasons (anastrozole: n = 9; fulvestrant: n = 4; A+F: n = 9) (Figure)

Table 3
details the determination of ESD status by treatment arm.The pCR rate was 0.5% to 1.2%, and the mPEPI 0 rate was 17.5% to 21.9% across the treatment arms.The ESDR was 18.7% with anastrozole, 22.8% with fulvestrant, and 20.5% with A+F (Table

Table 3 .
Neoadjuvant Endocrine Therapy (NET) Outcomes and Endocrine-Sensitive Disease (ESD) Rate (ESDR) by Treatment Arm Unable to determine whether mPEPI was 0 or nonzero due to missing information on pT stage, pN stage, or Ki67.Testing H0: difference in ESDR of 10% or less against Ha: difference in ESDR greater than 10%.Downloaded from jamanetwork.comby The University Of North Carolina Chapel Hill user on 01/19/2024 a b
Abbreviations: A, anastrozole; aOR, adjusted odds ratio; F, fulvestrant.aHodges-Lehmann estimate of the difference in 2 medians with corresponding distribution-free confidence interval (Moses) based on the Wilcoxon rank sum test.b