Palbociclib Combined with Fulvestrant in Premenopausal Women with Advanced Breast Cancer and Prior Progression on Endocrine Therapy: PALOMA‐3 Results

In the randomized, phase III PALOMA‐3 trial, palbociclib plus fulvestrant (6 goserelin) prolonged investigator‐assessed progression‐free survival compared with placebo plus fulvestrant (6 goserelin) in women with HR1/HER2‐ advanced breast cancer after prior progression on endocrine therapy. PALOMA‐3 was the first registrational study to include premenopausal women in this setting; in this article, results are described by menopausal status, with a focus on premenopausal women with prior endocrine‐resistant HR1/HER2‐ advanced breast cancer.

Abstract

Background.

The efficacy and safety of palbociclib, a cyclin‐dependent kinase 4/6 inhibitor, combined with fulvestrant and goserelin was assessed in premenopausal women with advanced breast cancer (ABC) who had progressed on prior endocrine therapy (ET).

Patients and Methods.

One hundred eight premenopausal endocrine‐refractory women ≥18 years with hormone receptor–positive (HR+)/human epidermal growth factor receptor 2–negative (HER2−) ABC were among 521 women randomized 2:1 (347:174) to fulvestrant (500 mg) ± goserelin with either palbociclib (125 mg/day orally, 3 weeks on, 1 week off) or placebo. This analysis assessed whether the overall tolerable safety profile and significant progression‐free survival (PFS) improvement extended to premenopausal women. Potential drug‐drug interactions (DDIs) and ovarian suppression with goserelin were assessed via plasma pharmacokinetics and biochemical analyses, respectively. (ClinicalTrials.gov identifier: NCT01942135)

Results.

Median PFS for premenopausal women in the palbociclib (n = 72) versus placebo arm (n = 36) was 9.5 versus 5.6 months, respectively (hazard ratio, 0.50, 95% confidence interval: 0.29–0.87), and consistent with the significant PFS improvement in the same arms for postmenopausal women. Any‐grade and grade ≤3 neutropenia, leukopenia, and infections were among the most frequent adverse events reported in the palbociclib arm with concurrent goserelin administration. Hormone concentrations were similar between treatment arms and confirmed sustained ovarian suppression. Clinically relevant DDIs were not observed.

Conclusion.

Palbociclib combined with fulvestrant and goserelin was an effective and well‐tolerated treatment for premenopausal women with prior endocrine‐resistant HR+/HER2− ABC. Inclusion of both premenopausal and postmenopausal women in pivotal combination ET trials facilitates access to novel drugs for young women and should be considered as a new standard for clinical trial design.

Implications for Practice.

PALOMA‐3, the first registrational study to include premenopausal women in a trial investigating a CDK4/6 inhibitor combined with endocrine therapy, has the largest premenopausal cohort reported in an endocrine‐resistant setting. In pretreated premenopausal women with hormone receptor–positive advanced breast cancer, palbociclib plus fulvestrant and goserelin (luteinizing hormone–releasing hormone [LHRH] agonist) treatment almost doubled median progression‐free survival (PFS) and significantly increased the objective response rate versus endocrine monotherapy, achieving results comparable to those reported for chemotherapy without apparently interfering with LHRH agonist‐induced ovarian suppression. The significant PFS gain and tolerable safety profile strongly support use of this regimen in premenopausal women with endocrine‐resistant disease who could possibly delay chemotherapy.

Introduction

Many women presenting with early breast cancer (BC) are premenopausal; upon relapse with advanced breast cancer (ABC), most are postmenopausal, predominantly as a result of cytotoxic therapy and extended endocrine therapy (ET) treatment [1], or via natural changes. Although younger women represent a low percentage of the patients diagnosed with breast cancer (20–34 years, 1.8%; 35–44 years, 8.9%) [2], presentation of early stage cancer at ≤40 years is prognostic in luminal breast cancer with the risk for relapse [3], [4]. Survival outcomes are also worse for women <40 years than for older age groups, notwithstanding intense treatment regimens [3].

De novo metastatic and relapsed breast cancer can also occur in peri/premenopausal women (hereafter referred to as premenopausal women) [5], [6]. Yet premenopausal women have generally been excluded from large registrational trials involving hormonal agents to assess hormone‐positive ABC; clinical data for premenopausal women remain remarkably limited to only a few small phase 2 trials [7], [8], [9].

Endocrine resistance continues to pose serious clinical challenges [10]. Sequential ET is generally the preferred treatment for women with hormone receptor–positive (HR+) human epidermal growth factor receptor 2–negative (HER2−) metastatic BC [11]. Premenopausal women can benefit from ovarian suppression as part of their ET. Options include oophorectomy, radiotherapy [12], or medical suppression [13] with a luteinizing hormone–releasing hormone (LHRH) agonist. Although tamoxifen is used as the first‐line ET for premenopausal women with HR+ BC (preferably with ovarian suppression [14]), aromatase inhibitors (AIs) plus LHRH agonists can be more effective [11], [15]. Fulvestrant, a pure estrogen receptor (ER) antagonist and selective ER degrader devoid of estrogenic effects [16], has not been standard treatment for premenopausal women because of limited data supporting its biological effects [17]. In a neoadjuvant study of 66 premenopausal women with ER‐positive (ER+) BC receiving 250 mg fulvestrant as monotherapy, fulvestrant did not significantly alter markers of hormone sensitivity (ER, progesterone receptor, and Ki67) [18] as it did in postmenopausal women [19]. Conversely, 750 mg fulvestrant did elicit significant changes in the same markers, similar to changes observed in premenopausal women treated with tamoxifen, suggesting a potential need for higher doses of fulvestrant in premenopausal women [20].

When used in combination with ET, palbociclib, a selective inhibitor of cyclin‐dependent kinases 4 and 6, resulted in a significant improvement in progression‐free survival (PFS) compared with ET alone in both treatment‐naive ABC (PALOMA‐1 trial: letrozole vs. letrozole + palbociclib [21]; PALOMA‐2 trial: letrozole + placebo vs. letrozole + palbociclib [22]) and in women pretreated for HR+/HER2− ABC (PALOMA‐3 trial: fulvestrant vs. fulvestrant + palbociclib [6], [23]).

In the randomized, phase 3 PALOMA‐3 trial, palbociclib plus fulvestrant (± goserelin) prolonged investigator‐assessed PFS compared with placebo plus fulvestrant (± goserelin) in women with HR+/HER2− ABC after prior progression on ET (median PFS 9.5 vs. 4.6 months, respectively, hazard ratio [HR], 0.46 [95% confidence interval {CI}: 0.36–0.59], two‐sided log‐rank p < .0001) [6]. PALOMA‐3 was the first registrational study to include premenopausal women in this setting; herein we describe the results by menopausal status, with a focus on premenopausal women with prior endocrine‐resistant HR+/HER2− ABC.

Materials and Methods

PALOMA‐3 Study Design and Patients

The design of the PALOMA‐3 phase 3, randomized, double‐blind, placebo‐controlled trial and definitions of the efficacy and safety parameters assessed have been described in detail elsewhere [6], [23]. Key eligibility criteria include women ≥18 years with HR+/HER2− ABC whose disease had progressed on prior ET. One previous line of chemotherapy in advanced disease was allowed. Women were defined as premenopausal if they did not meet the criteria for postmenopausal status, defined as age ≥60 years, age <60 years and amenorrhea for ≥12 consecutive months (excluding an alternative pathologic or physiologic cause), and serum estradiol (E2) and follicle‐stimulating hormone (FSH) concentrations within the laboratory reference range for postmenopausal women, documented bilateral oophorectomy, or medically confirmed ovarian failure.

Patients were randomized 2:1 to receive palbociclib (125 mg/day orally, 3 weeks on, 1 week off) plus fulvestrant (500 mg intramuscularly on day 1 and 15 of cycle 1 and once every 28‐day cycle thereafter) or placebo plus fulvestrant. Premenopausal patients were required to receive a LHRH agonist subcutaneously every 28 days starting ≥4 weeks before study treatment, and upon starting treatment any patients using a LHRH agonist other than goserelin were switched to goserelin. Randomization was stratified by menopausal status, visceral metastases, and sensitivity to prior hormonal therapy [6], [23].

Biochemical assessments were performed on blood samples collected from premenopausal women on day 15 of study treatment. Plasma E2 analysis was conducted by InVentiv Health (Burlington, MA, http://www.inventivhealth.com) using gas chromatography (GC)/tandem mass spectrometry (MS‐MS). The lower limit of quantification was 0.625 pg/mL. Luteinizing hormone (LH) and FSH were measured at the Royal Marsden Hospital by immunoradiometric assay (MG12151, IBL International, Hamburg, Germany, http://www.ibl-international.com; KIP0841, DIAsource, Ottignies‐Louvain‐la‐Neuve, Belgium, http://www.diasource-diagnostics.com). The kit‐reported detection limits were 0.2 and 0.1 mIU/mL, respectively.

Plasma pharmacokinetic (PK) samples were drawn predose on days 1 and 15 of cycles 1 and 2 and on day 1 of cycle 3 for the assessment of trough concentrations (C_trough) of palbociclib, fulvestrant, and goserelin (when applicable) in a subgroup of ∼40 patients included in an initial interim safety assessment. Additional PK samples for plasma C_trough of palbociclib were drawn on day 15 of cycles 1 and 2 from all remaining patients.

Statistical Analysis

The Kaplan‐Meier method [24] was applied to estimate median progression‐free survival (mPFS) and generate survival curves. A two‐sided unstratified log‐rank test was used to compare treatment arms by menopausal status and a one‐sided unstratified log‐rank test was used to compare treatment arms in subsets of premenopausal and postmenopausal patients ≤50 years as part of an exploratory analysis. The HR was estimated from the Cox proportional hazards regression model. Clinical benefit response (CBR) and objective response rate were evaluated and compared between treatment arms using a one‐sided exact test stratified by the presence of visceral metastases and sensitivity to prior hormonal therapy per randomization. A multivariate analysis was run to evaluate the relationship between baseline prognostic factors with PFS. The final explanatory variables for the model were acquired using a backward selection process with a 0.1 significance level required for retaining the effects in the model. Descriptive analysis was used to summarize maximum‐grade treatment‐emergent adverse events (AEs). Biochemistry data were summarized and Student's t tests (one for each hormone) were used to compare data between treatment arms without a multiplicity adjustment for these or other exploratory analyses. The analyses for the potential for drug‐drug interactions (DDIs) are described in detail in the appendix (supplemental text in online Appendix 1). Statistical analyses were performed using SAS® Version 9.2 (SAS Institute, Cary, NC, https://www.sas.com/en_us/home.html).

Before any study procedures were initiated, all patients provided written informed consent. Institutional review boards at participating centers approved all study‐related procedures, which were conducted in accordance with the International Conference on Harmonisation, the guidelines for Good Clinical Practice, and the Declaration of Helsinki. Study enrollment is closed and the trial met its primary endpoint at interim analysis. The overall survival follow‐up is in progress.

Results

Between October 7, 2013, and August 26, 2014, 521 patients were randomly assigned to receive palbociclib plus fulvestrant (347 patients) or placebo plus fulvestrant (174 patients) with or without goserelin (supplemental online Fig. 1A) [6].

Patient Demographics

A total of 108 (21%) women were premenopausal and 413 (79%) were postmenopausal. Overall, 42 (8%) women were ≤40 years, and 163 (31%) women were ≤50 years. Among premenopausal women, the mean age was 45 years; 83 (80%) of these patients were ≤50 years. Among postmenopausal women, 80 (19%) patients were aged ≤50 years. Demographic and baseline characteristics are shown by menopausal status in Table 1. Half of the premenopausal women had received both tamoxifen and AI before enrollment. Baseline characteristics between treatment arms within menopausal subgroups were well balanced, except more premenopausal women had two prior lines of therapy in the ABC setting in the palbociclib than in the placebo arm (31.9% and 19.4%, respectively). Baseline characteristics between premenopausal and postmenopausal women were also well balanced, although the ESR1 mutation rate by cfDNA was slightly lower in premenopausal women.

Table 1. Baseline patient and disease characteristics for premenopausal and postmenopausal patients randomly assigned in PALOMA‐3.

^aDue to rounding, some percentages may not total to exactly 100%.

^bAt least one target lesion ≥20 mm by conventional techniques or at least one target lesion >10 mm for spiral CT.

^cOne subject with negative duration included in <12 months category.

^dPremenopausal women had to have documented progression while on or within 12 months of completion of adjuvant therapy with tamoxifen whereas postmenopausal women had to have similarly progressed on an aromatase inhibitor.

^eValues are mutually exclusive. Not all most recent prior therapies are shown.

^fSubjects are counted for each treatment of metastatic disease (± neoadjuvant) received.

^gSensitivity to prior hormonal therapy was defined as either (a) documented clinical benefit (complete response, partial response, stable disease ≥24 weeks) to ≥1 prior hormonal therapy in the metastatic setting, or (b) ≥24 months of adjuvant hormonal therapy prior to recurrence.

^hNot all patients had cfDNA samples; the data represent a subset of patients only.

Abbreviations: cfDNA, circulating free DNA; CT, computed tomography; ER+, estrogen receptor–positive; ESR1, estrogen receptor 1; LHRH, luteinizing hormone–releasing hormone; PI3K, phosphoinositide 3‐kinase; PR+, progesterone receptor–positive; PR−, progesterone receptor–negative.

Efficacy

The data cutoff for this analysis was March 16, 2015, corresponding with the final primary efficacy analysis of PFS for the overall intent‐to‐treat population [6]. Investigator‐assessed mPFS for palbociclib plus fulvestrant versus placebo plus fulvestrant in the premenopausal subgroup was 9.5 versus 5.6 months, respectively (HR, 0.50 [95% CI: 0.29–0.87], two‐sided p = .013). In the postmenopausal subgroup, mPFS was 9.9 versus 3.9 months (HR, 0.45 [0.34–0.59], two‐sided p < .0001). Kaplan‐Meier survival curves for PFS are shown for premenopausal women in Fig. 1 [25].

Figure 1.

Investigator‐assessed PFS by treatment for the intent‐to‐treat subpopulations of premenopausal women (A), premenopausal women age ≤50 years (B), and postmenopausal women age ≤50 years (C). PFS was defined as the time from the date of randomization to the date or the first documentation of objective progression of disease or death due to any cause in the absence of documented progressive disease, whichever occurred first. PFS data were censored on the date of the last tumor assessment on study for patients who did not have objective tumor progression and who did not die while on study. CI was calculated based on the Brookmeyer and Crowley method [25].

Abbreviations: CI, confidence interval; NE, not estimable; PFS, progression‐free survival.

In the palbociclib arm versus the placebo arm, investigator‐assessed objective responses were observed in 25.0% (18/72) versus 11.1% (4/36) of premenopausal patients, respectively (odds ratio [OR], 3.06 [95% CI: 0.82–13.38], p = .057; Fig. 2) [26]. Significant improvement occurred with palbociclib plus fulvestrant versus placebo plus fulvestrant in investigator‐assessed CBR, which was observed in 69.4% (50/72) versus 44.4% (16/36) of premenopausal women (OR, 2.89 [95% CI: 1.15–7.34], p = .011).

Figure 2.

Investigator‐assessed confirmed objective response and clinical benefit rate in premenopausal women. ^aOne‐sided exact test stratified by the presence of visceral metastases and sensitivity to prior hormonal therapy per randomization. ^bCBR was CR or PR or stable disease ≥24 weeks. ^cCI was calculated using the exact (Clopper‐Pearson) method [26].

Abbreviations: CBR, clinical benefit response; CI, confidence interval; CR, complete response; OR, odds ratio; ORR, objective response rate; PR, partial response; SD, stable disease ≥24 week.

Of those premenopausal women who underwent subsequent chemotherapy after disease progression, 58.6% (17/29) had received palbociclib plus fulvestrant, whereas 78.3% (18/23) had received placebo plus fulvestrant. The median time to first chemotherapy treatment, relative to the date of randomization, was 120.0 (range, 37–354) days for women in the palbociclib arm and 74.5 (53–240) days for those in the placebo arm.

Exploratory Findings in Pre‐ and Postmenopausal Women ≤50 Years Old

Among premenopausal women aged ≤50 years (n = 83), mPFS was 9.5 months in the palbociclib arm and 5.6 months in the placebo arm (HR, 0.53 [95% CI: 0.28–0.99], one‐sided unstratified log‐rank test, p = .022; Fig. 1B). Among postmenopausal women ≤50 years (n = 80) in the palbociclib arm compared with the placebo arm, mPFS was 7.7 versus 4.5 months, respectively (HR, 0.49 [0.27–0.89], one‐sided unstratified log‐rank test, p = .008; Fig 1C).

Prognostic Factors

Important and favorable prognostic factors in the final Cox proportional hazards multivariate model included absence of visceral disease for both premenopausal and postmenopausal patient subpopulations and Asian ethnicity for premenopausal patients. The treatment effect of palbociclib plus fulvestrant seen in the primary analysis of PFS held when the important prognostic factors were simultaneously adjusted in the multivariate analyses for both premenopausal and postmenopausal patients (Table 2). When body mass index (BMI) was examined with the treatment arm in a separate multivariate model, lower values were more favorable for PFS.

Table 2. Multivariate analyses of the association of baselinea/prognostic factors with progression‐free survival.

Palbociclib arm, palbociclib + fulvestrant; placebo arm, placebo + fulvestrant.

^aBaseline factors that entered the model selection included visceral disease (yes vs. no), time from first diagnosis to relapse (≤24 months vs. >24 months vs. NA), prior treatment (AIs vs. non‐AIs), prior chemotherapy (yes vs. no), prior endocrine therapy (1 line vs. >1 line), and race (Asian vs. non‐Asian [white, black, and other]).

^bPremenopausal status is per randomization.

^cA hazard ratio <1 indicates a reduced hazard in the first category, whereas a hazard ratio >1 indicates a reduced hazard on the last category of the variable.

^dTwo‐sided p value, bold indicates significant at the threshold of p < .05.

Abbreviations: AI, aromatase inhibitor; CI, confidence interval; NA, not available.

Safety

Although all premenopausal women received goserelin concurrently with palbociclib plus fulvestrant, the safety profile was similar to that of postmenopausal women in terms of the type and frequency of AEs (all grades and grade 3–4), serious AEs (SAEs), the rate of dose reductions, and cycle delays. Dose interruptions and discontinuation rate due to AEs were also similar for the palbociclib arm between menopausal subgroups (premenopausal, 5.6% and postmenopausal, 4.7%; Table 3).

Table 3. Summary and listing of treatment‐emergent AEs (all causalities, all cycles) occurring in ≥10% patients in any treatment arm and dose modifications due to AEs by menopausal subgroup and treatment.

^aPercentages are calculated in reference to n, and values include data up to 28 days after the last dose of study drug.

^bEvents coded using the Medical Dictionary for Regulatory Activities (version 18.0) PTs, including clusters of PTs, and by maximum Common Terminology Criteria for Adverse Events grade.

^cAverage daily dose administered = (total dose administered) ÷ (total days on drug). Days on drug is defined as the total number of days on which the drug was actually administered.

^dClustered PTs: Anemia refers to any event having a PT equivalent to anemia or hematocrit decreased or hemoglobin decreased; infections is any event having a PT part of the system of organ class infections and infestations; leukopenia is any event having a PT equivalent to leukopenia or white blood cell count decreased; neutropenia is any event having a PT equivalent to neutropenia or neutrophil count decreased. Rash is any event having a PT equivalent to dermatitis or dermatitis acneiform or rash or rash erythematous or rash maculopapular or rash papular or rash pruritic; stomatitis is any event having a PT equivalent to aphthous stomatitis or cheilitis or glossitis or glossodynia or mouth ulceration or mucosal inflammation or oral pain or oropharyngeal discomfort or oropharyngeal pain or stomatitis; thrombocytopenia is any event having a PT equivalent to platelet count decreased or thrombocytopenia.

^eGrade 5 events: in the palbociclib plus fulvestrant arm of the study, two (2.8%) premenopausal women had disease progression, and one (1.4%) premenopausal woman with disease progression also had hepatic failure; among postmenopausal women in the palbociclib plus fulvestrant arm, one (0.4%) had disseminated intravascular coagulation and one (0.4%) experienced general or physical health deterioration; among postmenopausal women in the placebo arm of the study, one (0.7%) had acute respiratory distress, one (0.7%) had breast cancer, and one (0.7%) had a cerebral hemorrhage.

Abbreviations: —, no data; AE, adverse event; Gr, grade; n, subjects evaluable for AEs (i.e., includes all patients who received ≥1 dose of study treatment [palbociclib/placebo or fulvestrant]); PT, preferred term.

Biochemical Analysis

After 15 days of study treatment, there was no significant difference in the mean concentrations of LH, FSH, or plasma E2 between those premenopausal patients receiving palbociclib or not; all unadjusted p values from the Student's t tests were >.05 (Fig. 3). Mean E2 concentrations were consistent with those expected in the postmenopausal range. One patient in the palbociclib arm had an E2 value of 93.5 pg/mL, more than 3 times the upper limit of that group and not consistent with ovarian suppression. There was a statistically significant correlation between plasma E2 levels and BMI in both treatment arms (Spearman's rho = .44, p = .002 and .49, p = .02, respectively).

Figure 3.

Biochemical plasma analyses for premenopausal women on day 15 for LH (A), FSH (B), and E2 (C). ^aSix patients had E2 concentrations that were not considered valid (i.e., below the quantification limit of 1.25 pg/mL or a volume too small for reanalysis).

Abbreviations: CI, confidence interval; E2, estradiol; FSH, follicle‐stimulating hormone; IQR, interquartile range; LH, luteinizing hormone; max, maximum; min, minimum.

Drug‐Drug Interaction Assessment

Results of the 1‐way analysis of variance (ANOVA) and analysis of covariance (ANCOVA) analyses conducted to investigate the potential for DDIs among palbociclib, fulvestrant, and goserelin are shown in the table in supplemental online Appendix 1. The ratio of the adjusted geometric means (90% CI) for palbociclib from the final ANCOVA model and the within‐patient mean steady‐state concentration trough (C_troughSS) in the presence and absence of goserelin was 88.3% (78.6%–99.1%). The ratio of the adjusted geometric means (90% CI) for goserelin within‐patient mean C_troughSS in the presence and absence of palbociclib was 110% (54.2%–225%). For the ratio of the adjusted geometric means (90% CI) for palbociclib from the final ANCOVA model, the within‐patient mean C_troughSS in the presence and absence of fulvestrant was 128% (117%–140%). The ratio of the adjusted geometric means (90% CI) for fulvestrant within‐patient mean C_troughSS in the presence and absence of palbociclib was 122% (101%–147%).

Discussion

We report the results of the largest cohort of premenopausal women with prior endocrine‐resistant HR+/HER2− ABC ever studied. The study found that premenopausal patients who received palbociclib plus fulvestrant significantly benefited compared with patients treated with placebo plus fulvestrant, consistent with the benefit observed for postmenopausal women. In premenopausal and postmenopausal women aged ≤50 years, PFS favored the palbociclib arm compared with the placebo arm, regardless of menopausal status and LHRH agonist therapy.

Among all postmenopausal women, only 9 (2.2%) were ≤40 years and 71 (17.2%) were aged >40–50 years. The majority of these women possibly became postmenopausal as a result of previous chemotherapy for either primary or ABC, or ovarian suppression, a scenario more likely to occur in women >40 years whose ovarian function is inherently less resilient to such treatment [27]. Breast cancer incidence has been shown to peak in Asian women at a younger age compared with women from Western countries [28]. As expected, the proportion of Asian women with ABC who were premenopausal (41% [44/108]) was at least twofold higher than for postmenopausal women (15% [61/413]) in this study.

Until recently, premenopausal women with HR+ metastatic BC who are candidates for ET were mainly treated with tamoxifen, with or without an LHRH agonist. In a small study of 26 patients, median age 44 (range, 30–51) years, 250 mg fulvestrant (dose currently suboptimal) plus goserelin as first‐ to fourth‐line therapy was reported to provide promising activity [29]. As an estrogen receptor downregulator, fulvestrant, unlike tamoxifen, is not efficacious in premenopausal women unless it is administered in combination with ovarian suppression [30]. In addition, a number of small phase 2 studies have shown the efficacy of AI treatment concurrent with LHRH agonists [7], [8], [31]. The addition of a LHRH agonist to tamoxifen or AIs has been shown to improve efficacy in early BC as well as ABC [14], [15], [27], [32], [33]. PALOMA‐3 is the only phase 3 study to date to report outcomes data for fulvestrant 500 mg with ovarian suppression in premenopausal patients [6]. In premenopausal women in the control arm, mPFS with fulvestrant was 5.6 months. Accordingly, the fulvestrant label approved by the U.S. Food and Drug Administration was recently expanded to include premenopausal women who receive concurrent ovarian suppression [34].

Because of the low potential for a DDI between the protocol‐specified concomitant medications and palbociclib based on their metabolic pathways and ability to affect the activity of relevant metabolic enzymes, this study was designed with the intention of confirming the lack of clinically significant DDIs using a sparse PK sample collection (C_troughSS only) for each analyte. The magnitude of the ratios of the adjusted geometric means for the ANOVA and ANCOVA DDI analyses was not considered to represent a clinically meaningful difference.

The PK data confirm there were no clinically significant metabolic DDIs between palbociclib and goserelin or between palbociclib and fulvestrant when these two drugs were coadministered. Furthermore, the coadministration of goserelin did not have a clinically significant impact on fulvestrant plasma PK.

The clinical effectiveness of LHRH agonists depends on their suppression of ovarian steroidogenesis. There is evidence that although these agents achieve complete cessation of ovarian E2 synthesis by the end of the first month of treatment, partial recovery can occur in some patients, apparently driven by progressive recovery of FSH levels [35]. Combining a LHRH agonist with fulvestrant could still potentially result in incomplete ovarian suppression if E2 and FSH levels increase in response to fulvestrant [35]. We report hormonal concentrations measured after 15 days of fulvestrant treatment but after ≥6 weeks on goserelin, given that goserelin was started 4 weeks before the study commenced. It is also important to note that the levels of E2 were measured with a highly sensitive GC/MS‐MS, which avoids cross reactions with fulvestrant or its metabolites that can lead to highly aberrant values that more commonly occur using immunoassays [36], [37]. Only one of 48 patients (2%) treated with palbociclib plus fulvestrant had an E2 value that was clearly premenopausal; otherwise, the values assessed for all patients were consistent with full ovarian function suppression, and there was no difference associated with the addition of palbociclib. This is supported by the significant relationship between plasma E2 levels and BMI; extragonadal estrogen production occurs mainly in subcutaneous fat [38] and resulting plasma concentrations are known to correlate with BMI [39]. Accordingly, there was no disruption of the correlation with the addition of palbociclib treatment. This underlines that the effect of palbociclib is independent of the background ET and supports the rationale for the concurrent use of LHRH agonists with ET in other studies currently recruiting participants [40], [41].

From a safety perspective, the incidence of any‐grade and grade 3–4 AEs and SAEs was similar between premenopausal and postmenopausal women who received palbociclib plus fulvestrant, despite the addition of goserelin to the regimen. Dose modifications of palbociclib were also similar between premenopausal and postmenopausal groups.

Conclusion

The palbociclib plus fulvestrant regimen, with the addition of goserelin, essentially enables premenopausal women to be treated in close accordance with the guidelines for postmenopausal women, as recommended by the National Comprehensive Cancer Network [30]. These findings show that the addition of palbociclib had no impact on the concentration of fulvestrant, complete ovarian suppression was maintained, and no additional toxicities were evident with the addition of a LHRH agonist, all factors important for the investigation of palbociclib in early stage breast cancer. The results support the use of palbociclib in combination with fulvestrant and goserelin for women with HR+ ABC, and these findings have expanded the treatment options for premenopausal women.

See http://www.TheOncologist.com for supplemental material available online.

Author Contributions

Conception/design: Nicholas C. Turner, Jungsil Ro, Massimo Cristofanilli, Hiroji Iwata, Nadia Harbeck, Sunil Verma, Justin Hoffman, Ke Zhang, Cynthia Huang Bartlett, Mitchell Dowsett

Provision of study material or patients: Hiroji Iwata, Seock‐Ah Im, Nadia Harbeck

Collection and/or assembly of data: Sibylle Loibl, Jungsil Ro, Hiroji Iwata, Seock‐Ah Im, Norikazu Masuda, Sherene Loi, Nadia Harbeck, Sunil Verma, Elizabeth Folkerd, Kathy Puyana Theall, Justin Hoffman, Ke Zhang

Data analysis and interpretation: Sibylle Loibl, Nicholas C. Turner, Jungsil Ro, Massimo Cristofanilli, Hiroji Iwata, Seock‐Ah Im, Norikazu Masuda, Sherene Loi, Fabrice André, Nadia Harbeck, Sunil Verma, Kathy Puyana Theall, Justin Hoffman, Ke Zhang, Mitchell Dowsett

Manuscript writing: Sibylle Loibl, Nicholas C. Turner, Jungsil Ro, Massimo Cristofanilli, Hiroji Iwata, Seock‐Ah Im, Norikazu Masuda, Sherene Loi, Fabrice André, Nadia Harbeck, Sunil Verma, Elizabeth Folkerd, Kathy Puyana Theall, Justin Hoffman, Ke Zhang, Cynthia Huang Bartlett, Mitchell Dowsett

Final approval of manuscript: Sibylle Loibl, Nicholas C. Turner, Jungsil Ro, Massimo Cristofanilli, Hiroji Iwata, Seock‐Ah Im, Norikazu Masuda, Sherene Loi, Fabrice André, Nadia Harbeck, Sunil Verma, Elizabeth Folkerd, Kathy Puyana Theall, Justin Hoffman, Ke Zhang, Cynthia Huang Bartlett, Mitchell Dowsett

Disclosures

Nicolas C. Turner: Pfizer (C/A, RF); Massimo Cristofanilli: Vortex‐consultant Dompe (C/A), Pfizer (H); Sherene Loi: Pfizer (RF); Fabrice Andre: Pfizer, Novartis, AstraZeneca, Roche (RF); Nadia Harbeck: Lilly, Novartis, Pfizer (C/A), Novartis, Pfizer (H); Sunil Verma: Novartis, Pfizer, Astra Zeneca, Eli Lilly, Roche (C/A, SAB), OncologyEducation.com (OI); Kathy Puyana Theall: Pfizer (E, OI); Justin Hoffman: Pfizer (E); Ke Zhang: Pfizer (E, OI); Cynthia Huang Bartlett: Pfizer (E); Mitch Dowsett: Pfizer (RF); The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board