Phase 2 randomized trial of primary endocrine therapy versus chemotherapy in postmenopausal patients with estrogen receptor-positive breast cancer

Abstract

BACKGROUND.

Few studies have compared primary neoadjuvant endocrine therapy with neoadjuvant chemotherapy in breast cancer patients. The need for preoperative chemotherapy with doxorubicin or taxanes may be reduced in postmenopausal patients with estrogen receptor (ER)-positive and/or progesterone receptor (PgR)-positive tumors. This randomized, controlled, phase 2 study evaluated the efficacy of neoadjuvant chemotherapy compared with endocrine treatment with aromatase inhibitors in postmenopausal women with ER-positive and/or PgR-positive breast cancer.

METHODS.

Eligible patients were randomly assigned to receive neoadjuvant anastrozole 1 mg/day (n = 61) or exemestane 25 mg/day (n = 60) for 3 months or doxorubicin 60 mg/m² with paclitaxel 200 mg/m² (four 3-week cycles). Study end points included overall objective response determined by palpation, mammography, and ultrasound, and the number of patients who qualified for breast-conserving surgery and radiotherapy.

RESULTS.

Clinical objective response was 64% in the endocrine therapy and chemotherapy treatment groups. Median time to clinical response was 57 and 51 days with aromatase inhibitors and chemotherapy, respectively (P > .05). Rates of pathological complete response (3% vs 6%) and disease progression (9% vs 9%) did not differ significantly in the endocrine therapy or chemotherapy group, respectively (P > .05). Rates of breast-conserving surgery were slightly higher in the endocrine group (33% vs 24%; P = .058). The most frequent toxicities from chemotherapy were alopecia (79%), grade 3/4 neutropenia (33%), and grade 2 neuropathy (30%). Endocrine treatment was well tolerated. No deaths occurred during the preoperative treatment.

CONCLUSIONS.

Preoperative neoadjuvant endocrine therapy with aromatase inhibitors was well tolerated and resulted in rates similar to chemotherapy in overall objective response and breast-conserving surgery in postmenopausal women with ER-positive and/or PgR-positive tumors. Cancer 2007. © 2007 American Cancer Society.

Preoperative neoadjuvant chemotherapy with agents such as doxorubicin and taxanes is an effective treatment for patients with breast cancer and leads to an increased rate of successful breast-conserving surgery and a decreased proportion of patients with metastatic involvement of the axillary lymph nodes.1 Neoadjuvant chemotherapy also provides an opportunity to assess potential responses of the tumor to a given agent, which is an important consideration in selecting postoperative (adjuvant) therapy. Data from large phase 2 and phase 3 chemotherapy trials have shown that 3 to 4 months of preoperative treatment can be given without compromising either locoregional control or long-term survival.2–6

Some early nonrandomized and randomized trials suggested that primary systemic chemotherapy might result in improved disease-free survival rates compared with standard adjuvant treatment,4, 5 but some of these trials were not designed as a direct comparison of preoperative and postoperative chemotherapy. In 1998, the National Surgical Adjuvant Breast and Bowel Project (NSABP) reported the result of a large prospective randomized trial (Protocol B-18) that compared 4 cycles of doxorubicin and cyclophosphamide (AC) given preoperatively to the same dose of AC given postoperatively.3, 7 The disease-free survival and overall survival rates for the 2 treatment arms of this trial were almost identical. B-18 demonstrated that clinical and pathologic tumor response were predictors of overall survival.

Similar to other reports, despite a 36% clinically complete response (cCR) rate, only 13% of all patients had a pathologically complete response (pCR), defined as the absence of invasive tumor in the breast.3, 7 A meta-analysis of 9 randomized studies (not involving taxanes) demonstrated the equivalence of neoadjuvant and adjuvant treatments for breast cancer in terms of survival, disease progression, and distant recurrence and showed that an increased risk of locoregional disease recurrence is associated with neoadjuvant treatment, especially when primary systemic treatment is not accompanied by any surgical intervention (eg, radiation therapy alone).8

The NSABP Protocol B-27 was designed to determine the effect of adding docetaxel after 4 cycles of preoperative doxorubicin and cyclophosphamide on clinical and pathological response rates and on disease-free survival and overall survival of women with operable breast cancer.9 Addition of docetaxel to doxorubicin and cyclophosphamide did not significantly affect disease-free survival or overall survival. There were trends toward improved disease-free survival with the addition of docetaxel. Preoperative docetaxel, but not postoperative docetaxel, significantly improved disease-free survival in patients who had a clinical partial response after doxorubicin and cyclophosphamide. Pathologic complete response, which was doubled (from 13% to 26%) with preoperative docetaxel, was a significant predictor of overall survival regardless of treatment.9

However, preoperative chemotherapy may be less effective in postmenopausal patients with estrogen receptor (ER)-positive and/or progesterone receptor (PgR)-positive tumors, at least with respect to doxorubicin-containing or taxanes-containing regimens.10, 11 pCR rates after chemotherapy were significantly higher among patients with tumors that were both ER negative and PgR negative compared with patients whose tumors had any (even low) expression of steroid hormone receptors.10 In the European Cooperative Trial in Operable Breast Cancer, pCR after neoadjuvant chemotherapy was observed in 42% of women with ER-negative tumors, compared with 12% in the ER-positive group.11, 12

In the NSABP B-27 study, however, it appears that patients with both ER-positive and ER-negative tumors demonstrated an increase in pCR rates with the addition of docetaxel. Although the proportional increase in pCR rates was similar in ER-positive and ER-negative tumors, ER-negative tumors had higher rates of pCR than did ER-positive tumors when treated with neoadjuvant AC, as well as when treated with doxorubicin and cyclophosphamide followed by docetaxel.9

This observation agrees with multiple studies in the neoadjuvant setting, thus indicating that ER- negative tumors are more sensitive to chemotherapy than are ER-positive tumors.10–12

Although neoadjuvant endocrine therapy has not been studied as extensively as neoadjuvant chemotherapy, the results to date have been comparable, with the notable exception of the breast-conserving surgery rate, presumably because of older patient populations enrolled in studies that used endocrine therapy.13 Several phase 2 studies have demonstrated the feasibility of various endocrine treatments in elderly breast cancer patients with advanced locoregional disease, thus suggesting an alternative approach to initial surgery.14, 15 Choice of patients for this strategy follows from well-established predictive and prognostic variables for response to endocrine therapy in more advanced disease, specifically the presence of ER and PgR, advanced age, and well-differentiated tumors. Most studies have reported lower objective response and pCR rates (50%–70% and <5%–10%, respectively) than have been reported for primary chemotherapy, and the time spent on endocrine therapy required to halve the tumor volume is often longer than that with chemotherapeutic agents (>3 months).16

The interesting report of a cohort study of 94 patients with operable breast cancer by Cameron et al., in which the choice of induction treatment (chemotherapy or endocrine therapy) was based on ER status, found no difference in overall survival between women given initial endocrine therapy and those given chemotherapy.14 Moreover, the contribution of pathologic axillary nodal status after primary treatment in predicting patients' outcomes was the same, regardless of primary treatment (endocrine therapy or chemotherapy). Nonetheless, observations favor primary endocrine treatment (alone or combined with chemotherapy) in patients whose tumors are rich in steroid receptors. Other investigators treated postmenopausal women aged younger than 70 years with ER-positive and/or PgR-positive operable breast tumors ranging from T2 >3 cm to T4 with tamoxifenbefore locoregional therapy. After a median follow-up of 83 months, median time with metastasis-free survival was 88 months, and overall survival was 102 months in women with T4 tumors.17 Three randomized controlled studies have evaluated the role of primary endocrine treatment in comparison with initial radiotherapy or surgery.15, 18, 19 None of these studies showed a difference for overall survival between the different treatments, although a nonsignificant delay in metastatic events when tamoxifen was given as induction treatment was reported in 1 study.19 Of note, the trial with the longest follow-up (>5 years) showed an increased failure of locoregional control when tamoxifen was given initially (61% vs 31%; P < .0001), but the trial included smaller tumors (<5 cm) than the other 2 studies, making a direct comparison difficult.19

To avoid chemotherapy-related toxicities, preoperative endocrine therapy is an alternative option for women with ER-positive or PgR-positive tumors.1, 20 For elderly postmenopausal women with comorbid conditions and large hormone-responsive tumors, breast-conserving surgery may be possible after well-tolerated, preoperative hormonal therapy. For these elderly patients, 4 months of letrozole resulted in improved tumor response and rates of breast-conserving surgery over tamoxifen.16, 21 The clinical response rate with presurgical letrozole was 60%, and 48% of patients underwent breast-conserving surgery. In contrast, the response rate with presurgical tamoxifen was 41%, with 36% of patients undergoing breast-conserving surgery. The rate of clinical progression during the 4 months of therapy was 8% in the letrozole arm and 12% in the tamoxifen arm.16, 21

Because a reduction in tumor volume may allow breast-conserving surgery and because of the potential for reduced efficacy of preoperative doxorubicin or taxanes in postmenopausal patients with ER-positive and/or PgR-positive tumors, alternative neoadjuvant therapies are under investigation. Before the current trial, there were few, if any, direct comparisons of primary neoadjuvant endocrine therapy with primary neoadjuvant chemotherapy in patients with hormone-responsive breast cancer. The objective of this study was to compare effects of neoadjuvant chemotherapy with neoadjuvant aromatase inhibitors on tumor responses and subsequent eligibility for breast-conserving surgery in postmenopausal women with ER-positive and/or PgR-positive breast cancer.

MATERIALS AND METHODS

Study Design

This was an open-label, randomized, phase 2 trial of once-daily endocrine therapy (exemestane or anastrozole) or chemotherapy (doxorubicin and paclitaxel, every 3 weeks for 4 cycles) in postmenopausal women with primary ER-positive breast cancer. The trial design is illustrated in Figure 1. All patients provided written informed consent, and institutional review board approval was obtained at each center.

Figure 1. Study design. This was a phase 2, open-label study. Postmenopausal breast cancer patients with ER-positive (ER+) and/or PgR-positive (PgR+) primary invasive cancer received either preoperative (primary) chemotherapy or endocrine therapy. Surgery was scheduled 3 months from the date the patient received her first treatment.

Eligible patients were randomly assigned 1:1 to receive endocrine therapy or chemotherapy according to a predetermined, computer-generated, randomized list that used permuted blocks. The same method of randomization was then used to assign patients within the endocrine group to receive exemestane or anastrozole. After they were randomly selected, patients received either exemestane 25 mg or anastrozole 1 mg daily for 3 months or combined chemotherapy (doxorubicin 60 mg/m² and paclitaxel 200 mg/m² every 3 weeks, 4 cycles) before surgery, unless the patient was withdrawn earlier because of progressive disease, an adverse event, or patient and/or investigator request. Surgery was scheduled 3 months from the date the patient received her first treatment. Patients who did not progress were considered to have completed the study when they had received 3 months of endocrine treatment or 4 cycles of chemotherapy and were assessed for surgery.

Treatment of each patient after surgery was at the investigator's discretion. It was typically recommended that patients receive tamoxifen for 5 years. All patients are being followed for development of local recurrence, distant metastases, and survival at 5 years postsurgery.

Study Population

Eligible patients were postmenopausal women with untreated, histologically confirmed, primary invasive breast cancer, with ER-positive and/or PgR-positive tumors and a life expectancy of at least 6 months. All tumors were stage IIA (T2, N0, M0) to IIIB (T4, N0-2, M0) without distant metastases defined as follows: tumor >2 cm but ≤5 cm in greatest dimension (T2), tumor >5.0 cm in greatest dimension (T3), tumor of any size with direct extension to chest wall or skin (T4); no regional lymph node metastasis (N0), metastasis to movable ipsilateral axillary lymph nodes (N1), or metastases to ipsilateral axillary lymph nodes fixed or matted (N2); and no distant metastasis (M0; patients were stage IIA to IIIB; the staging system used for this study may be found in the American Joint Committee on Cancer Staging Manual, 6th Edition).21 Patients with T2 tumors were not eligible for breast-conserving surgery. Postmenopausal status was defined as an absence of spontaneous menses for at least 1 year in women older than 55 years of age; amenorrhea in women younger than 55 years and postmenopausal gonadotropin levels (lutropin and follitropin levels >40 IU/L) or postmenopausal estradiol levels (<5 ng/dL), or levels within the postmenopausal range for the local laboratory; or bilateral oophorectomy before the diagnosis of breast cancer. ER and PgR status was determined by immunohistochemistry (Allred score) with >10% positive malignant epithelial cells or by the dextran charcoal method. Tumor size was determined by clinical examination, mammography, and ultrasound.

All patients were required to have documented evidence of adequate bone marrow function (white blood cells >3.5 × 10⁹/L, platelets >100 × 10⁹/L, and hemoglobin >11.0 g/dL), renal function (creatinine <120 μmol/L), and hepatic function (bilirubin <25 μmol/L, and serum glutamic-oxaloacetic transaminase <60 U/L).

Exclusion criteria included prior chemotherapy or endocrine therapy, uncontrolled cardiac disease, bilateral or inflammatory breast cancer, distant metastases, and other malignant diseases (except cervical carcinoma treated in situ or adequately treated basal or squamous cell carcinoma of the skin). Coadministration of other cancer treatments was not permitted, and hormone replacement therapy must have been discontinued for at least 2 weeks before the start of the study.

Study Assessments

At the initial study visit, demographics and a complete medical history were recorded for each patient. Patients also underwent a physical examination and a review of current medical conditions, cancer diagnosis, and menopausal status. The clinical stage of the tumor was determined by measuring the primary breast lesion and axillary lymph nodes (if applicable) by physical examination. Mammography, breast ultrasound, and samples for ER and PgR status were also obtained. Core biopsies for the correlative science protocol were obtained before the start of trial medication.

Patients were assessed for tumor response according to World Health Organization (WHO) criteria by monthly clinical examination. Tumor response was also evaluated by ultrasound at the end of Months 2 and 3. If early progression was suspected, ultrasound was performed at the end of Month 1. At the final visit, ultrasound, mammography, and assessment for surgery were performed, and sampling for the correlative science protocol was repeated.

Evaluation of adverse events and concomitant medications and/or therapies were also assessed at each monthly visit. Adverse events were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.22 An electrocardiogram, physical examination, and measurement of left ventricular ejection fraction were performed at diagnosis and after each phase of chemotherapy. The measurements were determined by a multigated angiogram or echocardiography with the same technique used throughout the study for each individual patient.

Study End Points

The primary efficacy end point was objective response rate defined as the percentage of patients in each treatment arm with a complete response (CR) or a partial response (PR) as determined clinically by breast palpation (CR, regression 100%; PR, regression >50%). Response categories were CR, PR, no change, progressive disease, or not assessable/not evaluable (NA/NE). Palpable ipsilateral axillary lymph node involvement downgraded a clinically complete response (cCR) to a PR.

Secondary efficacy end points included the percentage of patients who underwent breast-conserving surgery and the response rate (CR + PR) determined by mammography and ultrasound at 3 months. The number of patients with a pathologically complete response (pCR) at surgery was also recorded.

Surgical breast and axillary node resection specimens were carefully evaluated for pathologic tumor response. Pathologic determinations were made by institutional pathologists who were provided with guidelines for tumor assessment before starting the trial. Patients without residual cancer in the breast were considered to have had a pCR. Patients whose residual cancer consisted of only noninvasive cancer were classified separately. Because these definitions did not consider histological nodal status, patients with noninvasive or no residual cancer could have had positive axillary lymph nodes.

Statistical Methods

Based on the primary efficacy end point (objective response rate determined by clinical palpation), it was calculated that a sample size of 239 patients would provide 80% power to detect the absence of a difference between the endocrine group and the chemotherapy group at a 5% significance level (2-sided), by using the Fleiss formula.23 Based on a review of early data on neoadjuvant treatment, a response rate of 62% was expected for the aromatase inhibitors and 63% for chemotherapy.16

Treatments were compared by using the Mantel-Haenszel chi-square test stratified by baseline tumor size (T2,>T2) and nodal involvement (N0,>N0). Secondary end points, including the proportion of patients in each treatment group with a complete response or part response assessed by ultrasound or mammography, as well as the number of patients in each treatment group who underwent breast-conserving surgery, were analyzed by using the same statistical test. An exploratory logistic regression analysis was conducted to examine the influence of baseline tumor size (T2, >T2), nodal involvement (N0, >N0), and age (<70 years, ≥70 years) on objective response rate as determined by clinical palpation and on the rate of breast-conserving surgery.

If a patient discontinued study treatment earlier than 3 months (±2 weeks) and had a last assessment of progressive disease, the earlier diagnosis of progressive disease was counted. If a patient discontinued study treatment earlier than 3 months (±2 weeks) for other reasons, then the final response was considered to be NA/NE for the analysis.

All patients with a diagnosis of breast cancer documented at study entry, who received at least 1 dose of study medication, and who were treated at a center compliant with Good Clinical Practice were included in the efficacy intent-to-treat population. Safety analyses included all patients who received at least 1 dose of study medication.

RESULTS

A total of 239 postmenopausal women with ER-positive and/or PgR-positive breast cancer (tumor classifications T2N1-2, T3N0-l, T4N0M0) were randomized to receive neoadjuvant chemotherapy (n = 118), or hormonal therapy with exemestane (n = 60), or anastrozole (n = 61). Six patients (2 endocrine therapy, 4 chemotherapy) who did not receive study medication were excluded from analysis. All patients were considered to be ineligible for breast-conserving surgery at enrollment. Baseline characteristics of both treatment groups were well balanced according to age, hormone receptor status, and disease stage. The majority (>75%) of patients were older than 60 years of age. All patients had hormone receptor-positive tumors. Fifty-eight percent of patients with ER-positive or PgR-positive tumors received endocrine therapy, and 53% of patients with both ER-positive and PgR-positive tumors received chemotherapy. Tumors were classified as T2/3 in 86% and 79% of patients randomized to endocrine therapy and chemotherapy, respectively, and/or N0/1 in 84% of patients in both groups (Table 1).

Table 1. Baseline Patient Characteristics

Characteristics	Neoadjuvant endocrine therapy, n=121	Neoadjuvant chemotherapy, n=118
	No. (%)	No. (%)
ER indicates estrogen receptor; PgR, progesterone receptor.
Age, y
Median	68	67
<55	11 (9)	15 (13)
≥55–<60	11 (9)	13 (11)
≥60–<65	36 (30)	43 (36)
≥65–<70	27 (22)	34 (29)
≥70–<75	17 (14)	10 (8)
≥75–<80	17 (14)	3 (3)
>80,	2 (2)	0
Receptor status
ER+/PgR+	70 (58)	63 (53)
ER+/PgR–	34 (28)	40 (34)
ER–/PgR+	17 (14)	15 (13)
Clinical stage
T2	59 (49)	51 (43)
T3	45 (37)	42 (36)
T4	17 (14)	25 (21)
N0	19 (16)	17 (14)
N1	81 (67)	81 (69)
N2	21 (17)	20 (17)
Disease stage
IIa	2 (1.6)	3 (2.5)
IIb	57 (47)	59 (50)
IIIa	41 (34)	36 (30)
IIIb	21 (17)	20 (17)

Representative tumor responses, using clinical examination and mammography according to WHO criteria, are shown in Figure 2. There was no statistically significant difference in overall objective response (CR + PR) between neoadjuvant chemotherapy and endocrine therapy. The primary efficacy end point (CR + PR determined by palpation) was similar in the endocrine group (anastrozole, 62%; exemestane, 67%) compared with chemotherapy (63%, P > .5; Table 2). The median time to clinical response was 57 days in patients who were receiving endocrine therapy and 51 days in patients who were receiving chemotherapy (P > .98). pCR in the primary breast tumor occurred in 6% and 3% of patients receiving chemotherapy and endocrine therapy, respectively (P > .05). The responses determined by mammography and ultrasound were similar for endocrine therapy and chemotherapy: 60% versus 63% (P > .5) and 40% versus 46% (P > .5), respectively (Table 3). A secondary efficacy end point was the percentage of patients who underwent breast-conserving surgery. There was a nonsignificant increase in the proportion of patients suitable for breast-conserving surgery in the endocrine therapy group compared with the chemotherapy group (33% vs 24%; P = .058).

Figure 2. Mammographic response: (A) partial or (B) complete.

Table 2. Primary Efficacy Results

Efficacy end point	Endocrine therapy, n=121	Chemotherapy, n=118	P
	No. (%)	No. (%)
Clinical response, palpation	78 (64.5)	75 (63.6)	>.5
Complete response	12 (10)	12 (10)	>.5
Partial response	66 (55)	63 (53)	>.5

Table 3. Secondary Efficacy Results

Efficacy end point, n (%)	Endocrine therapy n=121	Chemotherapy n=118	P
	No. (%)	No. (%)
Mammographic response	73 (60)	74 (63)	.500
Complete response	7 (6)	8 (7)	.640
Partial response	66 (55)	66 (56)	.960
Ultrasound response	49 (40)	55 (46)	.800
Complete response	4 (3)	5 (4)	.920
Partial response	45 (37)	50 (42)	.690
Breast-conserving surgery	40 (33)	28 (24)	.058

There was a trend toward higher overall rates of objective response and breast-conserving surgery among patients with tumors expressing high levels of ER (Allred score ≥6) in the endocrine therapy group compared with the chemotherapy group (43% vs 24%; P = .054; Table 4).

Table 4. Overall Objective Response in Patients With High Levels of Estrogen Receptor Expression*

Response	Endocrine therapy, n=70	Chemotherapy, n=63	P
	No. (%)	No. (%)
* High levels of estrogen receptor expression are defined as ≥6 Allred score or ≥120 fmol/g.
Clinical objective response	49 (70)	38 (60)	.068
Mammography	46 (66)	38 (60)	.088
Breast-conserving surgery	30 (43)	15 (24)	.054

After completing neoadjuvant treatment, 31 (13%) patients did not undergo surgical resection: 5.7% of patients who were receiving endocrine therapy and 7.2% of patients who were receiving chemotherapy. Twenty-two patients did not receive surgery because of disease progression. These patients were switched to the other study therapy: patients initially treated with endocrine therapy received chemotherapy, and patients treated with chemotherapy received endocrine therapy. Progressive disease was observed in 9% of patients who were receiving endocrine therapy and in 9% of patients who were receiving chemotherapy (P > .5). Stable disease was seen in 21% of patients who were receiving endocrine treatment and in 26% of patients who were receiving chemotherapy. Nine patients remained inoperable.

After a mean follow-up of <36 months, local recurrence was observed in 3.3% and 3.4% of patients who received endocrine therapy and chemotherapy, respectively.

The incidence of commonly reported adverse events was higher in patients who were receiving chemotherapy (Table 5). No serious adverse events were reported in patients who were receiving endocrine therapy. Six patients who were receiving chemotherapy experienced febrile neutropenia that led to treatment interruption. No deaths occurred during preoperative therapy.

Table 5. Adverse Events Related to Treatment

Event	Endocrine therapy, n=121	Chemotherapy, n=118
	No. (%)	No. (%)
NCI CTCAE indicates National Cancer Institute Common Terminology Criteria for Adverse Events, version 3; LVEF, left ventricular ejection fraction.
Neutropenia (grade 2–4)	0	51 (43)
Febrile neutropenia	0	6 (5)
Infection (grade 2–3)	0	2 (2)
Stomatitis (grade 3)	0	8 (7)
Diarrhea	0	8 (7)
Neuropathy
NCI CTCAE 2	0	35 (30)
NCI CTCAE 3	0	2 (2)
Alopecia	0	93 (79)
Cardiotoxicity (LVEF <50%)	0	8 (7)
Hot flushes (grade 2)	28 (23)	2 (2)
Fatigue (grade 2)	18 (15)	9 (8)
Vaginal bleeding	8 (7)	0
Arthralgia (grade 1–2)	8 (7)	2 (2)
Myalgia	6 (5)	2 (2)

DISCUSSION

Surgery is an essential component in the management of breast cancer. Breast-conserving surgery is preferred, but not all women are eligible. Neoadjuvant therapy may increase the proportion of women eligible for breast-conserving surgery by reducing tumor size, and chemotherapy and endocrine therapy with tamoxifen and aromatase inhibitors have been investigated in this setting.

Neoadjuvant therapy should not replace standard adjuvant therapy. Rather, it represents a reasonable alternative for women with palpable unicentric cancers who require mastectomy, and it is an excellent arena for clinical trials.

Aromatase inhibitors have shown superiority over tamoxifen in permitting breast-conserving surgery when used as preoperative therapy for 3 to 4 months.24–28 An additional advantage of neoadjuvant treatment is that it can be used as an in vivo test to assess tumor sensitivity to the treatment. Clinical data support the hypothesis that the chemotherapeutic sensitivity of primary tumors and micrometastases is similar.1, 29, 30 Therefore, failure of preoperative systemic treatment may be predictive of the response to the same therapy administered postoperatively.

Our previous relatively small study of neoadjuvant anastrozole versus chemotherapy (doxorubicin with paclitaxel) showed similar clinical objective response.31 Therefore, it was interesting to compare the steroidal aromatase inactivator exemestane with the nonsteroidal aromatase inhibitor anastrozole in the neoadjuvant setting against chemotherapy. Interim analysis of this trial showed similar objective response in patients who were receiving exemestane and in patients who were receiving anastrozole. It allowed us to review and to analyze dates on all patients who were receiving aromatase inhibitors in the endocrine therapy group.

This report describes the first randomized, open-label, clinical trial comparing preoperative endocrine treatment (the aromatase inhibitors exemestane and anastrozole) and chemotherapy (4 cycles of doxorubicin with paclitaxel). The results of this study showed that 3 months of preoperative endocrine therapy is as effective as preoperative chemotherapy in postmenopausal women with ER-positive and/or PgR-positive breast cancer. The similar times to response with both treatments indicate that the preoperative treatment period of 3 months was not biased in favor of either endocrine treatment or chemotherapy.

Comparison of the outcome of this study with results from other studies of preoperative chemotherapy alone is difficult because most preoperative chemotherapy studies have not reported data specifically for postmenopausal patients with ER-positive and/or PgR-positive tumors.7, 9 Preoperative chemotherapy is considered to have higher response rates (65%–78%) than preoperative hormonal therapy.4, 16 However, it is clear that the efficacy of preoperative chemotherapy may be reduced in postmenopausal patients with ER-positive and/or PgR-positive tumors, at least with respect to regimens employing doxorubicin plus taxanes.10, 11 The endocrine response rate of 64% observed in this study is, therefore, highly encouraging, particularly given the simplicity and low toxicity of the regimen.

Several recent studies support the use of aromatase inhibitors as neoadjuvant therapy for hormone-responsive breast cancer. For example, we recently reported the results of a study comparing the efficacy of exemestane and tamoxifen as neoadjuvant therapy.32 In that study, 151 postmenopausal women with ER-positive and/or PgR-positive breast cancer were randomly assigned to receive exemestane or tamoxifen for 3 months. Neoadjuvant treatment with exemestane significantly improved clinical objective response (76% vs 40%; P = .05) and the rate of breast-conserving surgery (37% vs 20%; P = .05), but it did not result in any significant differences in objective response as determined by mammogram or ultrasound.32 Thus, exemestane is more effective than tamoxifen as a neoadjuvant treatment option for postmenopausal women with ER-positive disease.

At least 6 relatively small studies (n = 11 to 38) have evaluated neoadjuvant therapy with exemestane.33–38 Dixon et al evaluated the effect of neoadjuvant exemestane in 13 postmenopausal women with ER-positive, operable, and locally advanced breast cancer.33 Exemestane was given for up to 3 months. Median tumor volume evaluated by clinical examination, mammography, and ultrasound was reduced by 86%, 84%, and 83%, respectively. After treatment, 10 patients had breast-conserving surgery with clear margins, and 2 underwent mastectomy. In a phase 2 study reported by Tubiana-Hulin et al, 38 postmenopausal women with ER-positive operable breast cancer received 4 to 5 months of neoadjuvant exemestane.38 Tumor response was evaluated by using National Cancer Institute Response Evaluation Criteria in Solid Tumors.39 Six percent of patients had a clinically complete response, 65% had a partial response, 24% had stable disease, and 45% had breast-conserving surgery. Taken together, these studies suggest that neoadjuvant exemestane is effective against hormone receptor-positive breast cancer.

To increase the pathologic response rate, the combination of exemestane with chemotherapy as neoadjuvant treatment was evaluated in 2 phase 1 studies.36, 37 Wolf et al treated 14 patients with locally advanced breast cancer with exemestane and increasing doses of epirubicin.37 Ten of 14 patients were evaluable: 2 patients had a complete response, 4 had a partial response, 3 had stable disease, and 1 had progressive disease. Six patients received breast-conserving surgery. Lichtenegger et al carried out a phase 1 study in 11 patients with locally advanced breast cancer with exemestane and increasing doses of docetaxel.36 These investigators reported that 78% of patients had a partial response and 22% of patients had stable disease. The pathologic response rates were high; 78% of patients achieved grade 1 responses, and 22% achieved grade 2 responses by staging criteria described by Chevallier et al.40 Because the combination was well tolerated with promising clinical responses, a phase 2 study has been initiated.

Adjuvant endocrine therapy will continue to be an important component of care for patients with hormone-responsive breast cancer. Sufficient evidence from clinical trials now exists that demonstrates the superior efficacy and improved tolerability of the third-generation aromatase inhibitors over both tamoxifen and other second-line therapies such as aminoglutethimide and megestrol acetate for metastatic desease.25–28, 41–46

Several large studies (Arimidex, Tamoxifen Alone or in Combination [ATAC], Breast International Group 1–98 trial [BIG 1–98], National Cancer Institute of Canada MA17 trial [MA17], Austrian Breast and Colorectal Cancer Study Group [ABCSG], and the Arimidex Nolvadex [ARNO 95] Studies) attest to the value of third-generation aromatase inhibitors in theadjuvant systemic therapy of postmenopausal women with endocrine-responsive early breast cancer,26, 47–50 and such therapy has been recommended as part of the standard of care in this patient group.49, 51

Neoadjuvant endocrine therapy also has a role in the treatment of breast cancer, particularly in postmenopausal women with hormone-responsive disease who are unable to tolerate toxicities associated with chemotherapy or who may be ineligible for immediate surgery.1, 17 In these patients, neoadjuvant therapy offers the promise of improving survival or enabling subsequent breast-conserving surgery. As in advanced breast cancer, the third-generation aromatase inhibitors have proven to be better tolerated than tamoxifen, with efficacy comparable to or superior to neoadjuvant therapy. Thus, the aromatase inhibitors should be strongly considered as alternatives to tamoxifen, which has long been the standard of care.

This trial has shown that preoperative endocrine therapy with aromatase inhibitors (anastrozole or exemestane) offers the same rate of overall objective response and breast-conserving surgery as chemotherapy in postmenopausal patients with ER-positive tumors. The frequency of adverse events was higher among patients who were receiving chemotherapy. Endocrine treatment was well tolerated. Preoperative endocrine therapy with aromatase inhibitors is a reasonable alternative to preoperative chemotherapy for postmenopausal women with ER-positive disease in clinical situations in which the low toxicity of the regimen is considered an advantage, for example, in women older than 70 years of age. Additional studies that assess long-term outcomes of neoadjuvant endocrine therapy with aromatase inhibitors are needed.

Acknowledgements

We thank Gerard P. Johnson, PhD, and Janet E. Stead, BM, BS, who provided editorial assistance on behalf of Pfizer Inc.