A Randomized Placebo Controlled Phase II Trial Evaluating Exemestane with or without Enzalutamide in Patients with Hormone Receptor–Positive Breast Cancer

Endocrine therapies (ET) that target the estrogen receptor (ER) signaling pathway, including the aromatase inhibitor, exemestane, play a critical role in the treatment of patients with hormone receptor–positive (HR⁺) breast cancer; however, most patients with advanced disease will develop resistance (1). Aromatase inhibitors block the conversion of androstenedione to estrogen, increasing the concentration of androgens that could then stimulate the androgen receptor (AR; refs. 2, 3), which is expressed in >75% of HR⁺ breast tumors (4, 5). The role of the AR in breast cancer is unclear; it has been associated with better patient outcomes in patients with ER-positive (ER⁺) disease (6, 7), but has also been associated with resistance to ETs (8). AR expression was correlated with lack of response in patients treated with tamoxifen in the advanced setting or with aromatase inhibitors in the neoadjuvant setting (9, 10). Patients with higher levels of the AR relative to the ER experienced shorter disease-free survival following adjuvant therapy compared with those who had a lower ratio (8).

Enzalutamide, a potent inhibitor of AR signaling, is approved to treat men with castration-resistant prostate cancer and metastatic hormone-sensitive prostate cancer (11–16). In preclinical breast cancer models, enzalutamide blocked both estrogen- and androgen-mediated growth of HR⁺ cells (8, 10). Together with epidemiologic data, these results suggest blocking both AR and ER signaling in patients with HR⁺ breast cancer could provide additional benefit beyond ER inhibition alone and may prevent resistance.

Enzalutamide is well tolerated in women with advanced breast cancer (17) and has demonstrated clinical activity in women with advanced AR⁺ triple-negative breast cancer (TNBC; ref. 18). However, clinical development of enzalutamide, and to our knowledge other AR inhibitors, has been limited in ER⁺ breast cancer and TNBC, largely because of the inability to definitively identify the patients most likely to benefit from AR-targeted therapy. For these agents to have clinical utility in breast cancer, predictive biomarkers are essential. Here, we report on a phase II study to evaluate the efficacy, safety, and predictive biomarkers of enzalutamide in combination with exemestane in women with HR⁺ breast cancer.

This ongoing phase II, randomized, double-blind, placebo-controlled study (NCT02007512) evaluated the efficacy and safety of enzalutamide 160 mg with exemestane 50 mg (the enzalutamide arm) versus placebo with exemestane 25 mg (the control arm) in patients with ER⁺ and/or progesterone receptor–positive (PgR⁺), HER2 normal, advanced breast cancer. Because of the observed interaction of exemestane with enzalutamide, a strong CYP3A4 inducer, doubling the dose of exemestane to 50 mg was necessary to restore the exposure observed with 25 mg (17). Two parallel cohorts were enrolled. Cohort 1 enrolled patients who had not previously received ET for advanced breast cancer. Cohort 2 enrolled patients who had progressed following one ET for advanced breast cancer. Patients were randomized 1:1 using an interactive web recognition system (IWRS) to the enzalutamide or control arm of each cohort. Stratification in cohort 1 was based on prior adjuvant or neoadjuvant ET (yes vs. no). If yes, patients were additionally stratified by prior aromatase inhibitor (yes vs. no) and hormone resistance, defined as disease recurrence within 24 months after initiating adjuvant hormone treatment (yes vs. no). Stratification in cohort 2 was based on prior aromatase inhibitor for advanced disease (yes vs. no) and hormone resistance, defined as disease progression within 24 weeks after initiating hormone treatment in the advanced setting (yes vs. no). Patients in the control arms were given the option to enroll in an open-label treatment period receiving enzalutamide with exemestane after disease progression. This study was approved by the institutional review board at each participating site and conducted according to the provisions of the Declaration of Helsinki and Good Clinical Practice guidelines of the International Conference on Harmonisation. All patients provided written informed consent before participating.

Eligible patients were ≥18 years of age, postmenopausal women with metastatic or locally advanced breast cancer that was ER⁺, PgR⁺, or both (defined as ≥1% of tumor nuclei staining for either receptor by IHC or Allred score of 2 or more) and HER2 normal (defined as 0 or 1+ by IHC or negative for HER2 amplification by ISH for 2+ IHC disease), with measurable disease or nonmeasurable bone and skin disease and Eastern Cooperative Oncology Group performance status (ECOG PS) ≤1; patients had no brain metastases or leptomeningeal disease, and no history of seizure or any condition that may predispose to seizure. Prior chemotherapy and prior hormonal treatment in the neoadjuvant or adjuvant setting and ≤1 prior chemotherapy for advanced disease were allowed. Tissue samples were required for biomarker analysis. Tissue from the primary diagnosis was preferred; tissue from recurrent and/or metastatic sites could also be submitted.

Radiographic assessments were performed by the investigators every 8 weeks for the first year and every 12 weeks thereafter until disease progression per RECIST v1.1. Radiographic assessment of soft-tissue disease was based on CT or MRI and assessment of bone disease was based on CT with bone windows or MRI in patients with suspected bone-only lesions. Bone assessments were to be obtained up to 12 weeks before randomization. Patients with bone-only nonmeasurable disease were to have a bone assessment performed at a maximum interval of every 6 months. Determination of response was made by the investigators using RECIST v1.1 for patients with measurable disease.

While the study was enrolling, the protocol was amended to include development and analysis of a genomic signature potentially predictive of response to enzalutamide. This signature was developed using RNA expression data from baseline tumor tissue from patients on this study. As the biomarker signature was developed using tumor samples from a subset of this study's patients, there was the potential for overfitting bias affecting the analysis. Given this possible bias, we felt that the signature should first be validated in another dataset and will, therefore, report results of this biomarker analysis in a subsequent manuscript. In addition, exploratory analyses of relevant biomarkers were performed using RNA expression data. Detailed methods are provided in the Supplementary Materials and Methods. Assessment of plasma pharmacokinetics for exemestane and enzalutamide included collection of predose samples for both study drugs and postdose samples for exemestane at weeks 5, 9, and 17.

The primary endpoint was progression-free survival (PFS) in the intent-to-treat (ITT) population of each cohort. Secondary endpoints included safety, best objective response rate (ORR), duration of response, and clinical benefit rate at 24 weeks (CBR24), defined as the proportion of patients with a best response of complete response (CR), partial response (PR), or stable disease lasting at least 24 weeks (confirmation of CR or PR was not required). Safety was assessed in all patients who received ≥1 dose of the study drug.

The planned enrollment was approximately 240 patients (120 patients per cohort) to provide at least 180 PFS events (90 events per cohort). The sample size was chosen to provide adequate precision for the preliminary assessment of the target HR of 0.67 for PFS in patients with AR-positive (AR⁺) breast cancer. With at least 55 PFS events in the AR⁺ subset, the 85% confidence interval (CI) for the HR will exclude 1 (0.45–0.99). With at least 35 PFS events in the AR-negative subset, the 85% CI for the HR is estimated to be 0.55–1.46 for a target HR of 0.9.

After data from the AR IHC assay were analyzed, it was determined that the IHC assay was not sensitive enough to be used as a companion diagnostic and its use was therefore discontinued.

Estimates of median PFS were determined using the Kaplan–Meier method. Two-sided log-rank tests stratified according to stratification factors by IWRS were used for between-group comparisons of PFS; stratified Cox regression models were applied to estimate HRs. PFS was evaluated in subgroups defined by baseline patient and disease characteristics, and unstratified Cox regression models were applied to estimate HRs. Two-sided unstratified χ² tests were used for between-group comparisons of CBR24 and ORR. Efficacy analyses were performed on the ITT population and subgroups based on gene expression (see Supplementary Materials and Methods). No adjustments for multiplicity were performed.

Upon request, and subject to certain criteria, conditions, and exceptions (see: https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information), Pfizer will provide access to individual deidentified participant data from Pfizer-sponsored global interventional clinical studies conducted for medicines, vaccines, and medical devices (1) for indications that have been approved in the United States and/or European Union or (2) in programs that have been terminated (i.e., development for all indications has been discontinued). Pfizer will also consider requests for the protocol, data dictionary, and statistical analysis plan. Data may be requested from Pfizer trials 24 months after study completion. The deidentified participant data will be made available to researchers whose proposals meet the research criteria and other conditions, and for which an exception does not apply, via a secure portal. To gain access, data requestors must enter into a data access agreement with Pfizer.

In cohort 1 (ET naïve), 127 patients were enrolled in the ITT population (enzalutamide arm, n = 63 and control arm, n = 64). One patient in each arm did not receive study drug. In cohort 2 (one prior ET), 120 patients were enrolled in the ITT population (enzalutamide arm, n = 60 and control arm, n = 60). All patients received study drug. At the data cut-off date of September 23, 2016, 106 patients (83.5%) in cohort 1 and 114 patients (95.0%) in cohort 2 had discontinued treatment (Fig. 1). In both cohorts, the primary reason for discontinuation was disease progression. Twenty-one patients from the control arm of cohort 1 and 12 patients from the control arm of cohort 2 enrolled in the open-label treatment period.

Baseline patient and disease characteristics are presented in Table 1. In cohort 1, baseline characteristics were generally similar between treatment arms. Most patients had received prior neoadjuvant/adjuvant hormonal therapy (66.1%) and/or neoadjuvant/adjuvant chemotherapy (56.7%). In cohort 2, baseline characteristics were well balanced between treatment arms, however, more patients had received prior neoadjuvant/adjuvant chemotherapy and/or neoadjuvant/adjuvant hormonal therapy in the control arm than in the enzalutamide arm.

In the ITT population of cohort 1, 42 (66.7%) patients had a PFS event in the enzalutamide arm and 48 (75.0%) had an event in the control arm. Although the median PFS was numerically longer in the enzalutamide arm [11.8 months (95% CI, 7.3–15.9)] than in the control arm [5.8 months (95% CI, 3.5–10.9)], the treatment effect was not significant [HR, 0.82 (95% CI, 0.54–1.26); P = 0.3631; Fig. 2A]. The HRs in most subgroups of patients favored the enzalutamide arm, although the number of patients in each subgroup was low (Supplementary Fig. S1A). There were no significant differences in CBR24 or ORR between the enzalutamide arm and the control arm (Table 2). Median duration of response in patients with CR or PR was 14.0 months (95% CI, 5.6–not reached) in the enzalutamide arm and 9.1 months (95% CI, 3.2–10.2) in the control arm. Median PFS in the open-label extension was 2.9 months (95% CI, 1.8–10.7).

In cohort 2, 44 (73.3%) patients had a PFS event in the enzalutamide arm and 50 (83.3%) had an event in the control arm; median PFS was similar in the enzalutamide and control arms [3.6 months (95% CI, 1.9–5.5) vs. 3.9 months (95% CI, 2.6–5.4), respectively; HR, 1.02 (95% CI, 0.66–1.59); P = 0.9212; Fig. 2B]. There were no subgroups that clearly favored the enzalutamide or control arms, although the number of patients in each subgroup was low (Supplementary Fig. S1B). There were no significant differences in CBR24 or ORR between the enzalutamide arm and the control arm (Table 2). Median PFS was not calculated for cohort 2 in the open-label extension because of the low number of patients.

AR expression by IHC was examined in the first 112 patients enrolled; of those, 110 patients had >0% nuclear staining and only 2 patients had no nuclear staining. As such, efficacy results in the subset of patients with positive nuclear AR staining were similar to the ITT population (data not shown). Therefore, in a post hoc analysis, we further evaluated the relationship between drug response and molecular targets (the ARs and ERs) by analyzing mRNA sequencing data. Of note, 80% of the tissue was from the breast, rather than distant metastatic sites. In cohort 1, the treatment effect of enzalutamide was dependent on AR levels (P_interaction = 0.0048). Patients with high AR levels, defined as above the median, in the enzalutamide arm had a longer median PFS compared with the control arm [14.0 vs. 5.5 months, respectively; HR, 0.42 (95% CI, 0.22–0.77); Fig. 3A], while patients with low AR levels in the enzalutamide arm had a shorter median PFS compared with the control arm [5.6 vs. 8.2 months, respectively; HR, 1.47 (95% CI, 0.79–2.73); Fig. 3A]. There were no differences based on AR levels in cohort 2 (P_interaction = 0.627; Fig. 3B).

In cohort 1, patients with low levels of ESR1, defined as below the median, in the enzalutamide arm had a longer median PFS than those in the control arm [11.8 vs. 3.8 months, respectively; HR, 0.52 (95% CI, 0.29–0.93); Fig. 3C]. Whereas in patients with high ESR1 levels, there was no difference in median PFS between treatment arms [HR, 1.09 (95% CI, 0.57–2.09)]. However, the statistical interaction between ESR1 levels and treatment effect was not significant (P = 0.127). In cohort 2, there was no PFS benefit from the addition of enzalutamide regardless of ESR1 levels [high ESR1: HR, 1.37 (95% CI, 0.70–2.65) and low ESR1: HR, 0.73 (95% CI, 0.39–1.38); Fig. 3D], potentially reflecting the fact that these patients had all previously relapsed following one hormone treatment for advanced disease.

AR and ESR1 expression levels were moderately correlated in both cohorts (R = 0.47 in cohort 1 and R = 0.37 in cohort 2; Supplementary Fig. S2). However, patients in the enzalutamide arm with both high AR levels and low ESR1 levels had an even greater reduced risk of progression or death compared with the control arm in cohort 1 [HR, 0.24 (95% CI, 0.10–0.60); P = 0.0011; Fig. 3E], with the median PFS extended by 10.2 months (from 3.8 in the control arm to 14.0 months in the enzalutamide arm). This pattern was not apparent in cohort 2 (Fig. 3F).

A total of 229 patients were included in the pharmacokinetics analysis (n = 114 who received exemestane 25 mg and n = 115 who received exemestane 50 mg with enzalutamide 160 mg). Mean plasma concentrations of exemestane 50 mg with enzalutamide 160 mg were comparable with those for exemestane 25 mg alone (Supplementary Fig. S3B). Enzalutamide and exemestane plasma concentrations were similar across weeks 5, 9, and 17 (Supplementary Fig. S3), suggesting they reached steady state by week 5.

In cohort 1, the median duration of exposure was 40.9 weeks in the enzalutamide arm and 25.7 weeks in the control arm. In cohort 2, the median duration of exposure was 10.2 weeks in the enzalutamide arm and 16.1 weeks in the control arm. In both cohorts, the incidence of patients with grade ≥3 adverse events and serious adverse events was greater in the enzalutamide arms than in the control arms (Table 3). The most frequently reported (occurring in ≥15% of patients) any grade adverse events in the enzalutamide arm of either cohort were fatigue, nausea, hot flushes, headache, and arthralgia; these were almost all grade 1/2 (Table 3). The most frequently reported (occurring in ≥2% of patients) grade ≥3 adverse events in the enzalutamide arm were hypertension and hypercalcemia in cohort 1, and anemia, hypercalcemia, fatigue, and headache in cohort 2 (Supplementary Table S1).

This phase II study is the first to report a randomized trial of enzalutamide in combination with exemestane for the treatment of HR⁺ breast cancer. In the unselected ITT population, there was no difference in PFS between patients receiving exemestane with enzalutamide or exemestane monotherapy. Adverse events were consistent with those reported in men with castration-resistant prostate cancer and in women with AR⁺ TNBC (11, 13, 18, 19). The lack of enzalutamide benefit observed in the ITT population of this study is consistent with a randomized phase II study of exemestane alone or in combination with the androgen biosynthesis inhibitor, abiraterone acetate, that also failed to show a significant benefit to anti-androgen therapy in an unselected population of patients with HR⁺ advanced breast cancer (20).

In our study, AR expression by IHC was evaluated as a potential predictive biomarker, however, it was not sensitive enough to be used as a companion diagnostic. In a phase II study of enzalutamide in TNBC, while AR expression by IHC was not associated with enzalutamide benefit (21), a gene signature–based biomarker indicating AR signaling was identified as potentially predictive of response to enzalutamide in the same study (22, 23). This is consistent with our finding that IHC was not sufficiently sensitive to be associated with enzalutamide benefit, but in the ET-naïve cohort, high AR mRNA levels were. Samples from our study were analyzed with the signature from the TNBC trial, however, the signature did not identify a population of enzalutamide-treated patients with HR⁺ breast cancer with better clinical outcomes.

We also evaluated other potential biomarkers on the basis of tumor biology. In the ET-naïve cohort, the subgroup of patients with low levels of ESR1 mRNA expression benefited from the addition of enzalutamide to exemestane. Interestingly, the combination of low ESR1 and high AR expression identified a population of patients that appeared to derive a particularly large benefit from the combination of enzalutamide and exemestane compared with exemestane alone [HR, 0.24 (95% CI, 0.10–0.60)]. This observation may suggest that these tumors are more dependent on AR signaling. In contrast, patients with high ESR1 levels and low AR levels perform better on exemestane alone, consistent with the notion that ER signaling is more functionally relevant in these patients; although, paradoxically, they have reduced benefit when enzalutamide is added to exemestane. The lack of association between exemestane benefit and expression of AR or ER in cohort 2 may reflect the fact that these were heavily pretreated patients who had progressed on prior ET. These patients' tumors may have developed resistance mechanisms reliant on alternative pathways and less dependent on AR or ER signaling. Taken together, these results point to a complicated interplay between ER and AR, as their relative ratio of expression appear to impact response to AR inhibition, which is also supportive of results reported in preclinical studies (24).

Our study is limited in that it was a small population treated before availability of CDK 4/6 inhibitors, and the biomarker analysis was exploratory and was based mainly on tumor samples from the breast, rather than distant metastatic lesions. In addition, the relatively small sample size required the biomarker analysis to be based on median cutoffs, rather than more granular categorizations, which potentially could be more biologically relevant. It is clear that the mechanism of AR signaling in breast cancer is complex and appears to be dependent on the presence or absence of other signaling mechanisms (25). The role of the AR in HR⁺ breast cancer, the predictive value of the identified biomarkers, and appropriate sequencing with other effective agents are still unclear and will require further studies.

I. Krop reports grants from Pfizer (to institution) during the conduct of the study, grants and personal fees from Genentech/Roche, and personal fees from Bristol Myers Squibb, Daiichi Sankyo, MacroGenics, Context Therapeutics, Taiho Oncology, Seattle Genetics, Novartis, Merck, Celltrion, and AstraZeneca outside the submitted work. V. Abramson reports personal fees from Eisai (consultant) and Daiichi Sankyo (consultant) outside the submitted work. M. Colleoni reports personal fees from Novartis (honoraria) outside the submitted work. T. Traina reports other from Pfizer (research support, honoraria), Astellas (research support), and Innocrin (research support, honoraria) during the conduct of the study, as well as other from Roche (research support, honoraria), Merck (honoraria), and AstraZeneca (research support, honoraria) outside the submitted work. L. Garcia-Estevez reports personal fees from Roche (lectures) and Eisai (lectures) outside the submitted work. L. Hart reports grants from Astellas Pharmaceuticals (to institution) and Pfizer (to institution) during the conduct of the study, as well as personal fees from Novartis, Genentech, Boehringer Ingelheim, and Lilly outside the submitted work. A. Awada reports grants from Roche, grants and personal fees from BMS, and personal fees from Lilly, ESAI, Pfizer, Novartis, Genomic Health, Ipsen, Bayer, Leo Pharma, Merck, Daiichi Sankyo, and Seattle Genetics outside the submitted work. C. Zamagni reports personal fees from Takeda, Pierre Fabre, Teva, and Istituto Gentili; grants, personal fees, and nonfinancial support from Roche, Novartis, AstraZeneca, Pfizer, Celgene, and Tesaro; personal fees and nonfinancial support from Eisai, PharmaMar, Lilly, Amgen, and QuintilesIMS; and grants from Roche/Genentech, Medivation, AbbVie, ArrayBiopharma, Morphotek, Synthon, Seattle Genetics, and Daiichi Sankyo outside the submitted work. P.G. Morris reports personal fees from Astellas, Pfizer, Teva, BMS, AstraZeneca, Genomic Health, and Roche, as well as personal fees and other from Novartis (consultant) outside the submitted work. L. Schwartzberg reports grants and personal fees from Amgen [consultant, research funding (institution)], as well as personal fees from Pfizer (consultant), Helsinn (consultant), Genentech (consultant), Genomic Health (advisory board), BMS (consultant), Myriad (consultant), AstraZeneca (consultant), Spectrum (consultant), Napo (consultant), and Bayer (DSMB) outside the submitted work. A. Gucalp reports other from Pfizer (Medivation/Astellas, institutional research grants/clinical trial funding) during the conduct of the study; A. Gucalp also reports personal fees and other from Pfizer (consulting/advisory board; institutional research grants/clinical trial funding) and Innocrin (travel funding and advisory board/steering committee; institutional research grants/clinical trial funding), and other from Merck (institutional research grants/clinical trial funding), Roche (institutional research grants/clinical trial funding), Novartis (institutional research grants/clinical trial funding), Oncotherapy (institutional research grants/clinical trial funding), BioAlta (institutional research grants/clinical trial funding), Zenith Epigenetics (institutional research grants/clinical trial funding), and Bristol Myers Squibb (institutional research grants/clinical trial funding) outside the submitted work. L. Biganzoli reports personal fees from AstraZeneca, Ipsen, Lilly, Pfizer, Daiichi-Sankyo, and Roche; grants and personal fees from Celgene and Genomic Health; and grants and personal fees from Novartis outside the submitted work. J. Tarazi reports other from Pfizer Pharmaceuticals (employment) during the conduct of the study. Z. Zhu reports employment with and ownership interest in Pfizer. E. Winer reports personal fees from Carrick Therapeutics (consultant), G1 Therapeutics (consultant), Genomic Health (consultant), GSK (consultant), Jounce (consultant), Leap (consultant/SAB member), Lilly (consultant), Novartis (consultant), Seattle Genetics (consultant), Syros (consultant), and Zymeworks (consultant), as well as grants and personal fees from Genentech/Roche (consultant and research to institute) outside the submitted work. No potential conflicts of interest were disclosed by the other authors.

I. Krop: Conceptualization, data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. V. Abramson: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. M. Colleoni: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. T. Traina: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. F. Holmes: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. L. Garcia-Estevez: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. L. Hart: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. A. Awada: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. C. Zamagni: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. P.G. Morris: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. L. Schwartzberg: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. S. Chan: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. A. Gucalp: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. L. Biganzoli: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. J. Steinberg: Conceptualization, data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. L. Sica: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. M. Trudeau: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. D. Markova: Conceptualization, data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. J. Tarazi: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. Z. Zhu: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. T. O'Brien: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. C.M. Kelly: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. E. Winer: Data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing. D.A. Yardley: Conceptualization, data curation, formal analysis, investigation, visualization, methodology, writing-original draft, writing-review and editing.

The authors would like to thank Iulia Cristina Tudor, Amy Peterson, and Hirdesh Uppal (former employees of Pfizer Inc.), Duncan Wheatley (Royal Cornwall Hospitals NHS Trust-Sunrise Centre, Cornwall, England), and Luca Gianni (Ospedale San Raffaele, Milan, Italy) for their contributions to the study and prior presentations. The authors would also like to thank Jarek Kostrowicki (Pfizer Inc.) for assistance with reviewing and analyzing the biomarker data. This study was funded by Pfizer Inc. and Astellas Pharma, Inc., the codevelopers of enzalutamide. The sponsors were responsible for the study design, data collection and analysis, decision to publish, and preparation of the article. Medical writing and editorial support funded by the sponsors were provided by Stephanie Vadasz, PhD, and Dena McWain of Ashfield Healthcare Communications.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.