IVT indicates intravaginal testosterone cream.
a7.5 μg/d (Estring 2 mg; Pfizer).
Overall, 34 patients completed treatment with intravaginal testosterone cream; 35 patients completed treatment with an estradiol-releasing vaginal ring (7.5 μg/d [Estring 2 mg; Pfizer]).
Overall, 34 patients completed treatment with intravaginal testosterone cream; 35 patients completed treatment with an estradiol-releasing vaginal ring (7.5 μg/d [Estring 2 mg; Pfizer]). IVT indicates intravaginal testosterone cream.
Overall, 34 patients completed treatment with intravaginal testosterone cream; 35 patients completed treatment with an estradiol-releasing vaginal ring (7.5 μg/d [Estring 2 mg; Pfizer]). CARES indicates Cancer Rehabilitation Evaluation System.
eTable 1. Demographics of 75 patients who initiated treatment
eTable 2. Clinical Characteristics of Patients with Baseline Elevated E2
eTable 3. Elevations in E2 (in pg/ml) as measured by LC/MS
eFigure 1. Total testosterone levels (in ng/dl) as measured by LC/MS in IVT patients
eFigure 2. CARES Sexual dysfunction and interest
Customize your JAMA Network experience by selecting one or more topics from the list below.
Melisko ME, Goldman ME, Hwang J, et al. Vaginal Testosterone Cream vs Estradiol Vaginal Ring for Vaginal Dryness or Decreased Libido in Women Receiving Aromatase Inhibitors for Early-Stage Breast Cancer: A Randomized Clinical Trial. JAMA Oncol. 2017;3(3):313–319. doi:https://doi.org/10.1001/jamaoncol.2016.3904
Does the use of an estradiol-releasing vaginal ring (7.5 μg/d) or intravaginal testosterone cream (IVT) lead to elevation in systemic estradiol in postmenopausal patients with breast cancer using aromatase inhibitors?
In this randomized, noncomparative trial, persistent estradiol elevation was observed in no patients using a vaginal ring and in a small portion of IVT patients. Vaginal atrophy, sexual interest, and sexual dysfunction improved in both groups.
Persistent estradiol elevation was rare among patients with breast cancer on aromatase inhibitors using a vaginal ring or IVT, and use of these products is reasonable to consider for patients experiencing urogenital atrophy.
Aromatase inhibitors (AI) are associated with significant urogenital atrophy, affecting quality of life and drug compliance.
To evaluate safety of intravaginal testosterone cream (IVT) or an estradiol-releasing vaginal ring (7.5 μg/d) in patients with early-stage breast cancer (BC) receiving an AI. Intervention was considered unsafe if more than 25% of patients had persistent elevation in estradiol (E2), defined as E2 greater than 10 pg/mL (to convert to pmol/L, multiply by 3.671) and at least 10 pg/mL above baseline after treatment initiation on 2 consecutive tests at least 2 weeks apart.
Design, Setting, and Participants
Postmenopausal (PM) women with hormone receptor (HR)–positive stage I to III BC taking AIs with self-reported vaginal dryness, dyspareunia, or decreased libido were randomized to 12 weeks of IVT or an estradiol vaginal ring. Estradiol was measured at baseline and weeks 4 and 12 using a commercially available liquid chromatography and tandem mass spectrometry assay; follicle-stimulating hormone levels were measured at baseline and week 4. Gynecologic examinations and sexual quality-of-life questionnaires were completed at baseline and week 12. This randomized noncomparative design allowed safety evaluation of 2 interventions concurrently in the same population of patients, reducing the possibility of E2 assay variability over time and between the 2 interventions.
Main Outcomes and Measures
The primary objective of this trial was to evaluate safety of IVT or an estradiol vaginal ring in patients with early-stage BC receiving an AI; secondary objectives included evaluation of adverse events, changes in sexual quality of life using the Cancer Rehabilitation Evaluation System sexuality subscales, changes in vaginal atrophy using a validated 4-point scale, and comparison of E2 levels.
Overall, 76 women signed consent (mean [range] age, 56 [37-78] years), 75 started treatment, and 69 completed 12 weeks of treatment. Mean (range) baseline E2 was 20 (<2 to 127) pg/mL. At baseline, E2 was above the postmenopausal range (>10 pg/mL) in 28 of 76 women (37%). Persistent E2 elevation was observed in none with a vaginal ring and in 4 of 34 women (12%) with IVT. Transient E2 elevation was seen in 4 of 35 (11%) with a vaginal ring and in 4 of 34 (12%) with IVT. Vaginal atrophy and sexual interest and dysfunction improved for all patients.
Conclusions and Relevance
In PM women with early-stage BC receiving AIs, treatment with a vaginal ring or IVT over 12 weeks met the primary safety end point. Baseline elevation in E2 was common and complicates this assessment. Vaginal atrophy, sexual interest, and sexual dysfunction were improved. Further study is required to understand E2 variability in this setting.
clinicaltrials.gov Identifier: NCT00698035
Aromatase inhibitors (AIs) are the preferred adjuvant therapy for postmenopausal (PM) women with hormone receptor–positive (HR+) early-stage breast cancer (BC).1,2 Vaginal dryness and decreased libido may lead to poor compliance and early discontinuation, and may adversely affect quality of life (QOL).3,4 Sexual dysfunction in patients taking AIs is common, with up to 50% reporting low sexual interest and 74% reporting insufficient lubrication.5
Vaginal estrogens are an effective treatment for urogenital atrophy, but safety is uncertain in patients with BC due to potential systemic absorption. Elevated estradiol (E2) was documented in patients taking AIs using the 25-µg Vagifem (Novo Nordisk) insert.6 Estring 2 mg (Pfizer), a 17 β-estradiol–releasing vaginal ring, reduces vaginal symptoms in PM women with little to no increase in E2.7 However, increases in E2 up to 60 days after insertion of Estring or the 25 µg Vagifem insert were observed in PM women at high risk for or with a history of HR+ breast cancer.8 Vaginal testosterone cream has also been shown to improve vaginal atrophy symptoms in PM women taking AIs.9
We conducted a prospective, randomized, open-label, single-institution phase 2 trial evaluating the safety and efficacy of intravaginal testosterone cream (IVT) or a vaginal ring in PM women with HR+ breast cancer taking AIs. Previous studies evaluating safety of intravaginal hormonal products have utilized research laboratory based radioimmunoassays (RIA) to measure E2 elevation.6,10 Because data exists suggesting that liquid chromatography and tandem mass spectrometry assays (LC/MS) are highly sensitive in this population,11,12 we compared baseline and week-4 E2 values using both RIA and LC/MS to determine if commercially available LC/MS could reliably detect E2 elevations.
The primary objective of this trial was to evaluate safety of IVT or a vaginal ring in patients with early-stage BC receiving an AI. Intervention was considered unsafe if more than 25% of patients had persistent elevation in E2, defined as E2 greater than 10 pg/mL (to convert to pmol/L, multiply by 3.671) and at least 10 pg/mL above baseline after treatment initiation on 2 consecutive tests at least 2 weeks apart. Definition of transient elevation in E2 was elevation without confirmation on consecutive blood draw with ongoing use of assigned product.
Secondary objectives included evaluation of adverse events, changes in sexual QOL using the Cancer Rehabilitation Evaluation System sexuality subscales,13 changes in vaginal atrophy using a validated 4-point scale,14,15 and comparison of E2 measured by a commercially available LC/MS assay or RIA.
The trial protocol is available in Supplement 1.
The study was approved by the University of California-San Francisco (UCSF) Committee on Human Research. All patients provided written informed consent. Eligible patients included PM women with stage I to III HR+ BC treated with at least 30 days of AI who also had vaginal dryness, dyspareunia, or decreased libido. Postmenopausal was defined as amenorrhea for at least 12 months in the absence of chemotherapy or bilateral oophorectomy. Women with chemotherapy-induced amenorrhea within 2 years of study entry, or premenopausal women receiving ovarian suppression and AI were eligible providing E2 was 10 pg/mL or less during the 30-day screening period. Exclusion criteria included history of vaginal or vulvar radiation, gynecologic cancer, abnormal pap smear test within 1 year, or concurrent estrogen or androgen treatment within 30 days. Initiation of topical moisturizers was prohibited, although continuation of previously used products was permitted.
Patients were randomized 1:1 to receive IVT or a vaginal ring for 12 weeks. Intravaginal testosterone cream was prepared by the UCSF Drug Product Services Laboratory as follows: micronized, United States Pharmacopeia (USP)-grade testosterone powder (synonym 4-Androsten-17β-ol-3-one) was levigated with USP-grade mineral oil to form a paste, mixed with a water-miscible odorless compounding vehicle (Velvachol), to produce a 1% IVT concentration. Patients were instructed to insert 0.5 mg of cream vaginally using a calibrated applicator daily for 2 weeks, then 0.5 mg 3 times a week for the remaining 10 weeks. After 24 patients were enrolled, the protocol was amended to reduce the frequency of IVT insertion to 3 times per week for patient convenience.
Estring 2 mg is a silicone ring for vaginal insertion that releases 7.5 μg of estradiol every 24 hours for 90 days. Participants or a study investigator inserted the ring on day 1.
A commercially available ultrasensitive LC/MS method was used to measure E2 at baseline, 4, and 12 weeks (Quest Diagnostics)16; those randomized to IVT also had serum ultrasensitive total testosterone levels measured (Quest Diagnostics).17 Patients were notified if their E2 was greater than 10 pg/mL and were given the opportunity to withdraw from study. Patients with E2 greater than 10 pg/mL at week 4 or 12, or with a greater than 10 pg/mL increase in E2 (if baseline was >10 pg/mL) had E2 drawn between weeks 6 to 8 (for week 4 elevations), or weeks 14 to 18 (for week 12 elevations). Baseline follicle-stimulating hormone levels were measured in a subset of patients.
Gynecologic examination was performed at baseline and week 12 or at end of study therapy. Vaginal epithelium was assessed using a 4-point scale evaluating rugae, pallor, petechiae, elasticity, and dryness. Changes were characterized as none, mild, moderate, or severe compared with normal epithelium.14,15 Patients completed the Cancer Rehabilitation Evaluation System sexual interest and sexual dysfunction subscales13 and single-item sexual satisfaction at baseline, week 4, and week 12.
Adverse events were monitored by telephone at week 4 and at the week 12 visit and graded using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 3.
This was an open-label, noncomparative phase 2 trial designed to test the null hypothesis that each treatment is unsafe (as defined in the Methods section) at a 5% level of significance.
Randomization was performed by the study coordinator in permuted blocks accounting for use of antidepressants and relationship status (single vs partnered) to balance for potential confounding factors. Each treatment arm was evaluated separately, with an early stopping rule if more than 25% of patients in either arm had persistent E2 elevation. With 35 patients in each arm, the trial was designed to have a 98% power of concluding treatment is safe if the true failure rate is 5% and 70% power when the true failure rate is 10%.
Changes between baseline and week 12 individual and composite vaginal atrophy scores were compared using standard t tests. Changes in the Cancer Rehabilitation Evaluation System sexual interest and dysfunction subscales, and single item sexual satisfaction were calculated for each arm. Adverse events and reasons for treatment discontinuation were collected at baseline and weeks 4 and 12.
Seventy-six patients were enrolled between October 2007 and June 2011. Sixty-nine patients (35 vaginal ring and 34 IVT) completed 12 weeks of treatment, and 68 patients (35 vaginal ring and 33 IVT) were evaluable for the primary end point (Figure 1). Baseline characteristics were well balanced (eTable 1 in Supplement 2). One patient randomized to the vaginal ring signed consent but did not start treatment. Four patients discontinued the vaginal ring (2 patients during week 3 for inability to maintain ring in vaginal vault, 1 patient during week 3 for vaginal odor/discomfort, and 1 patient during week 8 for a family emergency). Two patients discontinued IVT (1 patient for a herpes outbreak with vulvar cellulitis at week 3, and another patient due to concern related to increase in week-4 E2 to 20 pg/mL). One IVT patient was lost to follow-up after week 12 with E2 of 22 pg/mL.
At baseline, mean and median E2 were 20 pg/mL and 6 pg/mL (range, <2 to 127 pg/mL), respectively, and were similar in the 6 patients receiving ovarian suppression plus AI. Mean baseline E2 was 26.8 pg/mL (95% CI, 14.5-39.1 pg/mL) for the vaginal ring and 9.0 pg/mL (95% CI, 5.1-12.9 pg/mL) for IVT patients (P = .04). Baseline E2 was elevated in 37% of patients (19 of 40 [48%] with the vaginal ring and 9 of 36 [25%] with IVT; χ2P < .05); eTable 2 in Supplement 2 describes the characteristics of patients with baseline elevations. One patient who underwent oophorectomy 3 months prior to study entry had baseline E2 elevation (98 pg/mL) that fell to 10 pg/mL and remained in the PM range despite vaginal ring treatment. Baseline follicle-stimulating hormone levels were PM in 21 of 24 patients (88%) with baseline E2 elevation. Three patients with premenopausal follicle-stimulating hormone levels were receiving ovarian suppression plus AI and met eligibility criteria with a PM E2 (<10 pg/mL) during screening, with low-level E2 elevation on day 1 ranging from 11 pg/mL to 16 pg/mL. No association was found between baseline E2 and age, weight, AI type, or body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) for all enrolled patients.
Both treatment arms met the protocol defined primary safety endpoint. Figure 2 shows the individual values and mean E2 at each time point for patients completing 12 weeks of therapy. For the vaginal ring, 4 of 35 patients (11%) experienced a transient E2 elevation (range, 11-29 pg/mL), and none had persistent elevations. For IVT, 4 of 34 patients (12%) had persistent elevation (range, 16-45 pg/mL) and another 4 of 34 patients (12%) had transient E2 elevation (range, 11-113 pg/mL). Although there was no protocol defined plan to compare treatment arms, more IVT patients had persistent E2 elevation than patients using the vaginal ring (4 vs 0). One vaginal ring patient and 1 IVT patient had isolated week-4 E2 elevations. Two patients using the vaginal ring and 1 IVT patient had an elevated week-12 E2 that dropped at week 14 to 18 following treatment discontinuation. In 1 IVT patient, week 4 E2 was 23 pg/mL, fell to less than 2 pg/mL at week 6, but increased to 22 pg/mL at week 12 with continued IVT use. The patient was subsequently lost to follow-up. Estradiol levels in patients with a protocol defined elevation are detailed in eTable 3 in Supplement 2. There was no relationship between baseline E2 and subsequent elevations. Only 1 patient with baseline E2 elevation experienced a subsequent protocol-defined elevation—an IVT patient with baseline E2 of 31 pg/mL increasing to 113 pg/mL at week 4. She elected to remain on study; week 6 E2 was 11 pg/mL.
Mean and median baseline total testosterone levels measured in 27 IVT patients were 33 ng/dL (to convert to nmol/L, multiply by 0.0347) and 27 ng/dL, respectively. Mean and median testosterone levels increased to 186 ng/dL and 82 ng/dL at week 4 and then fell to 171 ng/dL and 67 ng/dL at week 12. Elevations in total testosterone outside the normal female PM range (2 ng/dL-45 ng/dl) were common with only 3 of 27 patients remaining in the PM range at both time points (eFigure 1 in Supplement 2).
Sixty-three patients had matched baseline and week-4 E2 by both assays. There were no differences between mean (SD) E2 with LC/MS and RIA at baseline (17.7 [28.5] pg/mL and 17.9 [44.1] pg/mL, respectively) or at week 4 (7.8 [15.0] pg/mL and 2.9 [13.4] pg/mL, respectively); baseline E2 was highly correlated between tests (Pearson r = 0.78; P < .001 and Spearman ρ = 0.43; P < .001). For patients with matched samples, baseline E2 was above PM in 9 of 63 patients (14%; all vaginal ring) by RIA, and in 25 of 63 patients (39%; 16 vaginal ring, 9 IVT) by LC/MS. Discordant E2 elevation was unusual; in 1 patient using the vaginal ring (baseline: 127 pg/mL by RIA vs 10 pg/mL by LC/MS; week 4: 107 pg/mL by RIA vs LC/MS <2 pg/mL), and in 1 IVT patient (baseline: 31 pg/mL by LC/MS vs 0.82 pg/mL by RIA; week 4: 113 pg/mL by LC/MS vs 0.82 pg/mL by RIA). Estradiol was elevated by LC/MS at week 4 in 19% of patients, compared with 1.5% by RIA.
Sixty-nine patients had baseline and week 12 vaginal atrophy evaluations (Figure 3). Individual and composite scores improved from baseline to week 12 for both vaginal ring and IVT. In patients using the vaginal ring, atrophy scores (lower score represents improvement) for rugae declined from 1.6 to 0.6, pallor from 1.4 to 0.6, petechiae from 1.1 to 0.1, elasticity from 0.9 to 0.3, and dryness from 1.5 to 0.5. Similar improvements were observed in IVT patients where rugae declined from 1.5 to 0.9, pallor from 1.5 to 0.7, petechiae from 1.1 to 0.5, elasticity from 1.2 to 0.4, and dryness from 1.5 to 0.8 (P < .001 for all comparisons, except petechiae in IVT which was .01).
Sixty-two patients completed all questionnaires (32 vaginal ring, 30 IVT); 6 patients completed a portion. For patients using the vaginal ring, mean (SD) sexual interest improved from 1.2 (0.9) at baseline to 0.9 (0.7) at week 12 (P = .021) and for IVT patients, from 1.4 (0.8) to 1.0 (0.6) (P = .02). Mean (SD) sexual dysfunction improved from 2.9 (1.1) at baseline to 2.0 (1.1) at week 12 in vaginal ring patients and from 2.9 (0.8) to 1.9 (1.1) in IVT patients (both P < .001). Mean (SD) single-item Cancer Rehabilitation Evaluation System sexual satisfaction score showed more improvement in vaginal ring patients, increasing from 2.5 (1.6) at baseline to 4.0 (1.5) at week 12 (P = .004) than in IVT patients, changing from 3.2 (1.6) to 4.0 (1.5) (P = .14) (Figure 4; eFigure 2 in Supplement 2).
All treatment-related adverse events were grade 1 or 2. Events occurring in more than 2% of participants included vaginal discharge (4 vaginal ring, 2 IVT [8% overall]), facial hair growth (1 vaginal ring, 5 IVT [8% overall]), vaginal or vulvar itching and/or irritation (4 vaginal ring [5% overall]), vaginal odor (3 IVT [4% overall]), and urinary tract or yeast infection (1 vaginal ring, 3 IVT [5% overall]). Two patients discontinued therapy for toxic effects (vaginal odor or irritation in vaginal ring and vulvar irritation in IVT).
To our knowledge, this is the largest study to date describing changes in E2, vaginal atrophy, and sexual quality of life in patients with BC using the vaginal ring or IVT for AI-related urogenital symptoms. Both the vaginal ring and IVT met the protocol-defined primary safety end point with 12% modest sustained E2 elevations with IVT compared with none with the vaginal ring.
Safety of vaginal testosterone for treating urogenital atrophy was previously evaluated in 21 patients with BC taking AIs.8 Using a lower dose of testosterone propionate (150-300 µg daily), no significant E2 elevation was reported at 4 weeks. The higher dose in our study (5000 µg 3 times/week) was based on data treating sexual desire disorders in PM women18 and resulted in testosterone levels above PM in most patients. While elevated testosterone levels may increase risk for BC,19 higher baseline testosterone levels were not associated with risk of recurrence in the WHEL study.20
Surprisingly, 37% of patients had elevated E2 at baseline by LC/MS. These results are intriguing, particularly given that 21% with elevated baseline E2 had undergone oophorectomy, and most had PM follicle-stimulating hormone levels. The reason for the significant difference in elevated baseline E2 between treatment arms is unclear because the arms were evenly paired for age, cause of menopause, BMI, race, and time on AI. Higher than expected E2 levels in AI-treated patients has been reported,21 although no large trial comparing AIs with tamoxifen in PM patients has measured baseline or longitudinal E2 levels.
Potential explanations for baseline E2 elevations include sporadic residual ovarian function or use of estrogen-containing supplements. Aromatase inhibitor compliance could have contributed to E2 variations, but this data was not formally collected. Although the protocol excluded treatment with oral, injectable, or topical form of estrogen or androgen therapy including natural supplements marketed as hormone replacement products, we did not specifically query on use of herbs. Use of dietary phytoestrogens and supplements including evening primrose, black cohosh, red clover extract, flaxseed, and wild yam is extremely common among patients with BC. Other less well-known products including epimedium, rehmannia, deer antler velvet, and cordyceps have estrogenic activity, and some have been demonstrated to contain E2 and increase serum E2 upon consumption.22-24 Lack of regulatory oversight and unclear labeling makes it difficult to monitor the content and safety of supplements and herbs, and contamination with natural or synthetic estrogens in complementary therapies has been documented.25 The presence of significant estrogen levels has been reported in youth-enhancing skin care products, none of which listed estrogenic molecules on their ingredient list.26
Affected patients were informed of their elevated baseline E2 and were advised to avoid supplements that might contain estrogen. All opted to continue study treatment. Subsequently, week-4 E2 was lower in 27 of 28 patients (96%) with baseline elevation, and only 5 of 28 (18%) had week-4 E2 above 10 pg/mL. The effect of sporadic E2 variations, use of herbs and/or supplements, and compliance will be even more important with increased AI use in premenopausal patients based on data from the SOFT trial.27 In the SOFT-EST substudy, over 17% of patients receiving triptorelin plus exemestane had an E2 above the PM threshold at 1 or more time points,27 demonstrating that chemical ovarian suppression is inconsistent.
We found that results from a commercially available ultrasensitive E2 assay (LC/MS) were largely concordant with those from an ultrasensitive RIA. Estradiol elevations were more frequently detected by LC/MS than RIA, where half were only modestly increased above PM range (14 of 28 of baseline E2 elevations by LC/MS were ≤20 pg/mL).
Both the vaginal ring and IVT improved vaginal atrophy, sexual interest, and desire, while only the vaginal ring improved sexual satisfaction. At the end of 12 weeks, patients had worse values for sexual interest, dysfunction, and satisfaction than previously reported in patients with BC several years out from acute treatment, many of whom were still on tamoxifen.28 This is likely owing to the greater effect of AIs on sexual function, as was observed in the SOFT27 and TEXT trials.29
The American Menopausal Society and American College of Obstetricians and Gynecologists recommend that in patients with a history of estrogen-sensitive cancers, nonhormonal options including vaginal moisturizers, pH balanced gels, topical oils, and lubricants be tried first for urogenital atrophy, and vaginal estrogens be reserved for patients who are nonresponsive and only after consultation with the medical oncologist.30,31
This study has several limitations, including relatively small size and the high percentage of patients with elevated baseline E2 levels before any intervention. Additionally, the conclusion that these interventions are safe based on lack of sustained E2 elevation may be considered arbitrary because there was no nonhormonal intervention for comparison and the study did not have statistical power or a long-term follow-up to explore whether there was a higher rate of recurrences in our study population vs historical controls.
This study provides further evidence that both an estradiol-releasing vaginal ring and IVT are effective in treating urogenital symptoms in patients with BC taking AIs. A small number of patients with sustained E2 elevations and high testosterone levels observed in the IVT arm suggest that lower-dose testosterone may be warranted. Elevated baseline E2 in patients who met all eligibility criteria was a surprising finding. Prospective data in a larger sample characterizing longitudinal E2 in PM women taking AIs and defining the effect of E2 fluctuation on outcome would be worthwhile. Careful attention should be paid to use of all supplements, including topical skin care products. For patients who experience symptomatic improvement on either an estradiol-releasing vaginal ring or IVT, our current standard clinical practice is to continue on therapy with measurements of serum E2 every few months depending on patient preferences.
Corresponding Author: Michelle E. Melisko, MD, Helen Diller Family Comprehensive Cancer Center, University of California-San Francisco, 1600 Divisadero St, Mt Zion Campus, Second Floor, San Francisco, CA 94115 (email@example.com).
Accepted for Publication: July 24, 2016.
Published Online: November 10, 2016. doi:10.1001/jamaoncol.2016.3904
Author Contributions: Dr Melisko had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Melisko, Goldman, Hwang, Esserman, Rugo.
Acquisition, analysis, or interpretation of data: Melisko, Goldman, Hwang, De Luca, Fang, Chien, Park, Rugo.
Drafting of the manuscript: Melisko, De Luca, Fang, Park, Rugo.
Critical revision of the manuscript for important intellectual content: Melisko, Goldman, Hwang, Esserman, Chien, Rugo.
Statistical analysis: Melisko, Hwang, Fang, Esserman.
Obtaining funding: Melisko.
Administrative, technical, or material support: Melisko, Goldman, De Luca, Fang, Chien, Rugo.
Study supervision: Melisko, Fang.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was funded in full by AstraZeneca.
Role of the Funder/Sponsor: AstraZeneca had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and decision to submit the manuscript for publication. AstraZeneca reviewed and approved the manuscript.
Previous Presentation: A portion of this study was presented at the 2009 San Antonio Breast Cancer Symposium; December 9 to 13, 2009; San Antonio, Texas.
Create a personal account or sign in to: