Paclitaxel After Doxorubicin Plus Cyclophosphamide As Adjuvant Chemotherapy for Node-Positive Breast Cancer: Results From NSABP B-28
http://www.100md.com
《临床肿瘤学》
the National Surgical Adjuvant Breast and Bowel Project, Operations Office and Biostatistical Center
The University of Pittsburgh Medical Center
Allegheny General Hospital, Pittsburgh, PA
Aultman Health Foundation, Canton, OH
Kaiser Permanente, Vallejo, CA
Rose Medical Center, Denver, CO
ABSTRACT
PURPOSE: The primary aim of National Surgical Adjuvant Breast and Bowel Project (NSABP) B-28 was to determine whether four cycles of adjuvant paclitaxel (PTX) after four cycles of adjuvant doxorubicin/cyclophosphamide (AC) will prolong disease-free survival (DFS) and overall survival (OS) compared with four cycles of AC alone in patients with resected operable breast cancer and histologically positive axillary nodes.
PATIENTS AND METHODS: Between August 1995 and May 1998, 3,060 patients were randomly assigned (AC, 1,529; AC followed by PTX [AC PTX], 1,531). Patients 50 years and those younger than 50 years with estrogen receptor (ER) or progesterone receptor (PR) -positive tumors also received tamoxifen for 5 years, starting with the first dose of AC. Postlumpectomy radiotherapy was mandated. Postmastectomy or regional radiotherapy was prohibited. Median follow-up is 64.6 months.
RESULTS: The addition of PTX to AC significantly reduced the hazard for DFS event by 17% (relative risk [RR], 0.83; 95% CI, 0.72 to 0.95; P = .006). Five-year DFS was 76% ± 2% for patients randomly assigned to AC PTX compared with 72% ± 2% for those randomly assigned to AC. Improvement in OS was small and not statistically significant (RR, 0.93; 95% CI, 0.78 to 1.12; P = .46). Five-year OS was 85% ± 2% for both groups. Subset analysis of the effect of paclitaxel according to hormone receptors or tamoxifen administration did not reveal statistically significant interaction (for DFS, P = .30 and P = .44, respectively). Toxicity with the AC PTX regimen was acceptable for the adjuvant setting.
CONCLUSION: The addition of PTX to AC resulted in significant improvement in DFS but no significant improvement in OS with acceptable toxicity. No significant interaction between treatment effect and receptor status or tamoxifen administration was observed.
INTRODUCTION
The value of adjuvant chemotherapy has been convincingly demonstrated in breast cancer patients with involved axillary nodes.1 Results from the most recent published update of the overview analysis indicate that administration of adjuvant chemotherapy significantly reduced the risk of recurrence by 23.5% and the risk of death by 15.3%. According to the same overview, the 10-year recurrence-free survival for node-positive patients treated with adjuvant chemotherapy was 47.6% (for patients younger than 50 years) and 43.6% (for those 50 to 69 years of age). The 10-year overall survival (OS) was 53.8% and 48.6%, respectively. Thus, despite significant progress in the treatment of node-positive breast cancer, there is still considerable room for improvement. In the overview analysis, the adjuvant chemotherapy most commonly used consisted of CMF (cyclophosphamide, methotrexate, and fluorouracil) or CMF-like regimens as well as anthracycline-containing regimens (administered for four or six cycles). When compared with patients treated with non–anthracycline-containing regimens, patients treated with anthracycline-containing regimens had a statistically significant reduction in recurrence rates (12%) and mortality rates (11%),1 although it has been argued that not all of the anthracycline-containing regimens included in the overview analysis are equally effective.2
The most commonly used anthracycline-containing adjuvant chemotherapy regimen in the United States consists of four cycles of doxorubicin plus cyclophosphamide (AC) administered every 21 days. This regimen became popular when results from a large adjuvant National Surgical Adjuvant Breast and Bowel Project (NSABP) study in node-positive patients who were deemed tamoxifen unresponsive (protocol B-15) showed its equivalence in efficacy and toxicity to the classic CMF regimen administered for six cycles (oral cyclophosphamide day 1 through 14 and methotrexate and fluorouracil on days 1 and 8).3 In addition, the AC regimen combined with tamoxifen was found to be superior to tamoxifen alone in another NSABP study conducted in patients who were deemed tamoxifen responsive (protocol B-16).4 Attempts to optimize the AC regimen by intensifying and/or increasing the dose of cyclophosphamide or that of doxorubicin did not result in improved outcome.5-7 One of the advantages of the AC regimen over classic CMF is that it can be given in much shorter time (84 days v 168 days) with fewer chemotherapy visits (four v 12). Thus, the AC regimen became attractive not only in the clinic but also in research studies as a control regimen to which one could sequentially add new non–cross-resistant drugs while keeping the total duration of adjuvant therapy to about 6 months.
In the early 1990s, the demonstration of significant antitumor activity with taxanes in patients with advanced breast cancer8-12 marked a new era in breast cancer chemotherapy and brought hope for the development of new, more active adjuvant chemotherapy regimens. Although the original clinical development of paclitaxel (PTX) started in the 1970s, initial progress was impeded by the occurrence of severe hypersensitivity reactions. In the late 1980s and early 1990s, hypersensitivity reactions were controlled with appropriate premedication and by administering the drug over longer periods of time (24 to 96 hours). Although longer infusions demonstrated significant clinical activity with good safety profile, they were impractical for the adjuvant setting. Thus, several studies explored and eventually confirmed the safety and efficacy of a 3-hour infusion at low or high doses,9,13-16 with 250 to 300 mg/m2 as a maximum-tolerated dose without colony-stimulating factor support.17 Of interest, a dose-response relationship with PTX had been originally postulated from data in ovarian cancer18 and was later also suggested in breast cancer studies,19,20 although this relationship was not as clear when intermediate and higher doses of PTX were compared with doses near the maximum-tolerated dose.20
In August of 1995, the NSABP initiated protocol B-28 in order to evaluate the worth of PTX as adjuvant chemotherapy for patients with resected operable breast cancer and histologically positive nodes. This randomized trial was designed to determine whether the addition of PTX (given at 225 mg/m2 as a 3-hour infusion) following four cycles of AC would prolong disease-free survival (DFS) and OS when compared with the administration of four cycles of AC alone (Fig 1). In this report, we present the first definitive analysis of this study.
PATIENTS AND METHODS
Patient Eligibility
Eligible patients were required to have resected operable adenocarcinoma confined to the breast and ipsilateral axilla on clinical examination (cT1-3, cN0-1, cM0) and to be randomly assigned within 63 days from initial cytologic or histologic diagnosis. Patients must have undergone either lumpectomy with free margins plus axillary node dissection or modified radical mastectomy, and the tumor had to be invasive adenocarcinoma with at least one positive axillary lymph node on pathologic examination. Determination of the estrogen receptor (ER) and progesterone receptor (PR) status of the primary tumor had to be performed before random assignment. Patients were required to have normal hematologic parameters and adequate hepatic and renal function as well as at least a 10-year life expectancy excluding their diagnosis of cancer. Eligible patients signed an approved consent form conforming to federal and institutional guidelines. Patients with a previous history of invasive breast cancer or ductal carcinoma-in-situ (in either breast) were ineligible, as were patients who had received any prior radiation, chemotherapy, immunotherapy, or hormonal therapy for their present breast cancer.
Protocol Therapy
Patients assigned to the control arm (AC) were to receive doxorubicin 60 mg/m2 by slow intravenous (IV) push followed by cyclophosphamide 600 mg/m2 IV infusion over 30 minutes to 2 hours every 21 days for four cycles. Patients assigned to AC followed by PTX [AC PTX] were to receive four cycles of AC as in the control arm, followed by four additional 21-day cycles of PTX 225 mg/m2 IV as a 3-hour infusion on day 1 of each cycle (AC PTX). Before each cycle of PTX, patients were required to receive premedication with dexamethasone 20 mg PO (12 and 6 hours before PTX infusion), diphenhydramine 50 mg IV (1 hour before PTX infusion) and cimetidine 300 mg IV or ranitidine 50 mg IV (1 hour before PTX infusion). All patients who were ER- or PR-positive, or who were 50 years of age or older at the time of surgery were to also receive tamoxifen 20 mg/day for 5 years, beginning on day 1 of the first AC cycle. Patients surgically treated with lumpectomy were required to receive whole-breast irradiation at a dose of 5,000 cGY in 25 fractions, with an optional external beam boost to the tumor site. Radiotherapy was administered following the last cycle of chemotherapy and after recovery from any toxicity. Postmastectomy local chest wall radiotherapy or regional nodal radiotherapy (irrespective of surgical procedure) was not permitted. Primary prophylaxis with colony-stimulating factors was not permitted, but secondary prophylaxis with recombinant human granulocyte colony-stimulating factor (rHu G-CSF; 5 mg/kg/day) was mandated following a cycle complicated by prolonged neutropenia, febrile neutropenia, or grade 3 to 4 infection. Antibiotic prophylaxis with ciprofloxacin was added to G-CSF prophylaxis following a cycle with grade 3 to 4 infection or after two cycles with febrile neutropenia. Dose reductions were required following a cycle with febrile neutropenia or grade 3 to 4 infection despite G-CSF and antibiotic prophylaxis. Dose reductions for PTX were required following a cycle with grade 3 neurotoxicity. PTX therapy was discontinued in case of grade 4 neurotoxicity, or in case of prolonged grade 2 or 3 neurotoxicity failing to return to grade 1 within 2 weeks. Similarly, PTX therapy was discontinued in case of severe hypersensitivity reaction or grade 3 to 4 cardiotoxicity.
Entry Studies and Follow-Up Requirements
Before assignment, patients were required to undergo history and physical examination (including gynecologic examination for patients with an intact uterus and/or ovaries), hematologic studies and chemistries, chest x-ray (within 3 months of random assignment), and ECG and bilateral mammogram (within 1 year from random assignment). Although a multi-gated pooled cardiac nuclear scan (MUGA) was not required, if performed it had to show a left ventricular ejection fraction of 45%. Patients were also required to have history and physical examination along with hematologic studies and chemistries on day 1 before each cycle of chemotherapy and every 6 months for the first 5 years. Gynecologic examination (when applicable), chest x-ray, and bilateral mammogram (unilateral for mastectomy patients) were required yearly for the first 5 years. Only physical examination, gynecologic examination (when applicable), and mammogram were required annually after 5 years.
Statistical Design and Methodology
Patient assignment to the two treatment arms was balanced with respect to histologic nodal status (one to three, four to nine, 10 positive nodes), assigned tamoxifen administration (no, yes), type of surgery (mastectomy, lumpectomy) and institution, using a biased-coin minimization algorithm.21 Participating sites faxed the required entry materials to the NSABP Biostatistical Center (Pittsburgh, PA), and random assignment was performed centrally. Patients' treatment assignments were then faxed back to the sites. The study was originally designed to have at least 80% power to detect a 25% reduction in mortality rate due to the sequential administration of PTX, using a .05-level two-sided statistical test for equality of mortality rates in the two study arms. When the trial was opened on August 1, 1995, the sample size was specified at 2,450 patients. After the review of early data indicating a lower percentage of patients completing all four cycles of PTX chemotherapy than was originally projected, the target sample size was increased to 3,050 patients. Definitive analysis was scheduled to take place after the report of the 490th death on both treatment arms combined. This ensures a power of 80% to detect a 22.6% reduction in mortality rate, in order to provide for a potentially attenuated treatment effect caused by noncompletion of therapy. Interim efficacy analyses were scheduled at protocol-specified intervals and took place after 64, 120, 269, and 369 deaths were reported. Early stopping boundaries were specified by the method of Fleming, Harrington, and O'Brien.22 In order to account for early interim looks, the final analysis is based on a .0452 two-sided level of significance. Following review of the third interim analysis, the Independent Data Monitoring Committee (DMC), which oversees all NSABP treatment trials, recommended that the interim results be reviewed at the National Institutes of Health Consensus Development Conference on Adjuvant Therapy for Breast Cancer (Bethesda, MD, November 1-3, 2000) even though the formal interim analysis boundaries had not been crossed. The DMC clearly expressed that this analysis should not be considered definitive. With this exception, no efficacy data were presented outside the DMC until the end point of 490 reported deaths was achieved.
Per protocol, the primary end points for analysis were DFS and OS. Events used in the determination of DFS include local, regional, and distant treatment failures, contralateral breast cancers, nonbreast second primary cancers, and deaths before recurrence or second primary cancer. The OS end point is based on deaths from any cause. Treatment comparisons of these two end points were based on log-rank tests, stratified by nodal status (one to three, four to nine, 10 positive nodes), protocol-assigned tamoxifen administration (no, yes) and operation (lumpectomy, mastectomy). Relative risks (RRs) and 95% CIs were estimated by fitting Cox models, with a dichotomous covariate representing assigned treatment arm and stratified in the same way. The analyses reported here are based on the intent-to-treat principle, and thus include all patients with follow-up, whether eligible or not, and regardless of the treatment actually received. Secondary analyses based on the eligible cohort of patients with follow-up gave substantively identical results. As an additional diagnostic check, treatments were compared after controlling for additional prognostic variables (number of positive nodes, clinical tumor size, operation, hormone receptor status, age, tamoxifen treatment, tumor grade, histologic tumor type) by fitting Cox models.
Based on post hoc findings reported in the Cancer and Leukemia Group B (CALGB) 9344 study,7 we also investigated whether the effectiveness of sequential AC PTX, measured on the basis of hazard ratios relative to treatment with AC alone, differed in hormone receptor (HR) -negative when compared with HR-positive patients. These tests were repeated for both the DFS and OS end points. Patients were considered to be HR-positive if they were either ER positive or PR positive. Hazard ratios were estimated separately for HR-negative and HR-positive cohorts by fitting Cox models after stratification for nodal status, tamoxifen, and operation. Log-hazard ratios were then compared using a Wald test to provide a test of PTX-by-HR interaction. A similar procedure was used to test for treatment-by-tamoxifen interaction. As an exploratory exercise, Cox models were also fitted to DFS and OS to screen for other interactions between treatment effect and prognostic factors (nodal status, clinical tumor size, operation, age at surgery, tumor grade, histologic tumor type). Analyses of recurrence-free interval (time to first recurrence censoring contralateral breast cancers, nonbreast second primary cancers, and deaths before diagnosis of cancer) were also performed. This end point was not specified in the protocol.
The individual end points defining first events (local, regional, distant recurrences, contralateral breast cancers, nonbreast second primary cancers, deaths before diagnosis of cancer) were analyzed by constructing a site-of-first-event table summarizing the frequencies of first events by treatment arm. For each end point, 5-year cumulative incidences were computed using the usual nonparametric estimator,23 and crude hazards were compared across treatment arms using the log-rank test.
Analyses reported here include all follow-up received at the NSABP Biostatistical Center as of the closure of the March 31, 2003, summary file, at which time 498 deaths and 863 DFS events had been reported. Median follow-up at that time was 64.6 months and was well balanced across treatment arms (Table 1). As of this report, vital status after 5 years is known for approximately 79% of all accrued patients. Only one patient has contributed no follow-up.
RESULTS
Accrual, Eligibility, and Patient Characteristics
The trial opened in August of 1995 and closed in May of 1998 after a accruing a total of 3,060 patients (1,529 patients were randomly assigned to AC and 1,531 to AC PXT). Thirteen patients in the AC arm and 11 patients in the AC PTX arm have been declared ineligible for various reasons (Table 1). Of 1,529 patients randomly assigned to AC, 1,528 (99.9%) had positive follow-up, compared with all of the 1,531 (100%) patients randomly assigned to AC PTX. Median follow-up is 64.8 months in the AC arm and 64.4 months in the AC PTX arm. Vital status at 5 years is known for 78.7% of patients in the AC arm and 79.9% of patients in the AC PTX arm.
Patient and tumor characteristics were distributed evenly between the two groups (Table 2). About half of the patients were younger than 50 years, and 19% were 60 years of age or older. Approximately 59% of the patients had tumors 2 cm in maximum diameter. Seventy percent of the patients in each arm had one to three positive nodes on pathologic examination, and only 4% of the patients had 10 positive nodes. Slightly less than half of the patients were treated with lumpectomy. About two thirds of the patients had ER-positive tumors, and approximately 61% to 62% had PR-positive tumors. As per protocol design, 85% of the patients were assigned to tamoxifen.
DFS and OS
There were 463 DFS events in the AC arm and 400 in the AC PTX arm (Table 3). Analysis of DFS demonstrated that the addition of PTX significantly reduced the risk of a DFS event by 17% (RR, 0.83; 95% CI, 0.72 to 0.95; P = .006). The 5-year DFS for patients in the AC arm was 72% ± 2% compared with 76% ± 2% for those in the AC PTX arm (Fig 2A).
There were 255 deaths in the AC arm and 243 in the AC PTX arm (Table 3). Analysis of OS demonstrated a non–statistically significant 7% reduction in death rate with the addition of PTX (RR, 0.93; 95% CI, 0.78 to 1.12; P = .46). The 5-year OS was 85% ± 2% for patients in both arms (Fig 2B). Although it appears that the curves are starting to separate after year 6, the number of patients at risk at the sixth and seventh year is relatively small, making any comparisons between the two arms at those time points inadvisable.
Table 4 shows the site of failure for all first events tabulated by treatment arm and summarizes crude hazard rates and 5-year cumulative incidence rates at each site. In addition to reducing recurrence rates, Table 4 suggests a possible effect of PTX in reducing the incidence of contralateral breast cancer (P = .039). There was also some reduction in the diagnosis of nonbreast second primary cancers in the AC PTX arm, although this effect was not statistically significant (P = .11). The cumulative incidence of all first events was 28.3% for the AC arm and 24.4% for the AC PTX arm.
Treatment effects for DFS and OS were also estimated by fitting multivariate Cox models that control for nodal status, clinical tumor size, operation, HR status, age, treatment with tamoxifen, histologic grade, and histologic tumor type. Estimates of treatment effect were almost identical to the ones obtained by stratified log-rank tests. For DFS, RR = 0.82 (95% CI, 0.72 to 0.94; P = .005; Table 5). For OS, RR = 0.92 (95% CI, 0.77 to 1.10; P = .34; data not shown). All of the prognostic factors were statistically significant in both the DFS and OS models, with the exception of histologic tumor type, which was only marginally significant for DFS (P = .09) and OS (P = .09). Prognostic factors were taken as reported and were not subjected to central review. There was no evidence of interaction between treatment effect and any of the factors, except for clinical tumor size. Treatment-by-tumor-size interaction was significant for OS (P = .02) but not for DFS (P = .08). However, the nature of the interaction suggested by the data appeared to be unlikely: PTX appeared relatively beneficial for either small tumors ( 2 cm) or large tumors (> 4 cm), but not for intermediate size tumors. This pattern of interaction appears to be spurious.
Effect of Treatment by HRs
Based on previous observation from CALGB 9344 relative to the effect of PTX according HR status, we undertook a similar subset analysis in this study. Analysis of the effect of PTX on DFS and OS according to HR status demonstrated no significant interaction between the effect of PTX in HR-positive compared with HR-negative patients (interaction P value .30 for DFS and .82 for OS; Table 3). The relative reduction in DFS events was 23% for HR-positive patients (RR, 0.77; 95% CI, 0.65 to 0.92; P = .004) and 10% for HR-negative patients (RR, 0.90; 95% CI, 0.72 to 1.12; P = .33; Table 3). The relative reduction in deaths was 6% for HR-positive patients (RR, 0.94; 95% CI, 0.74 to 1.21, P = .64) and 10% for HR-negative patients (RR, 0.90; 95% CI, 0.70 to 1.17, P = .44; Table 3).
We also performed a subset analysis of the effect of PTX on recurrence-free interval according to HR-status. Here again there was no evidence of an interaction relative to the effect of PTX in HR-positive compared with HR-negative patients. The relative reduction in recurrence was 12% for HR-positive patients (RR, 0.88; 95% CI, 0.72 to 1.07; P = .19) and 17% for the HR-negative patients (RR, 0.83; 95% CI, 0.65 to 1.05; P = .11) with an interaction P value of .69. Notice that although the interaction did not approach statistical significance, its sense was generally consistent with the data from CALGB 9344.
In addition to the above analyses, we examined the effect of PTX on DFS and OS according to tamoxifen administration (data not shown). In these subset analyses, there was no evidence of a significant treatment-by tamoxifen interaction.
Chemotherapy Completion Rate, Nonprotocol Therapy, and Toxicity
Chemotherapy completion rate was high during AC chemotherapy, with more than 98% of the patients completing all four cycles of AC. One hundred thirty-four patients of the 1,531 randomly assigned to AC PTX (8.8%) did not start PTX; 88% of those were due to patient withdrawal as opposed to physician withdrawal. The chemotherapy completion rate dropped during administration of PTX with 75.9% of the patients completing all eight cycles of protocol chemotherapy. For patients in the AC PTX arm in which treatment with PTX began but was discontinued before completion of eight cycles of therapy, physician withdrawal was the reason in 53% of the cases, with patient withdrawal accounting for the other 47%.
Toxicity from protocol therapy was acceptable for the adjuvant setting. There were seven deaths where treatment could not be excluded as a contributing factor. Five occurred in patients who received AC only (pulmonary embolism in one, congestive heart failure in two, sepsis in one, and seizure in one) and two in patients who received AC and PTX (coronary artery disease in one, pulmonary embolism in one). Most common grade 3 or greater toxicity during PTX therapy (based on the highest toxicity grade reported during PTX cycles) included neurosensory toxicity in 15% of patients, neuromotor toxicity in 7%, (grade 3 or higher neurosensory or neuromotor toxicity in 18%), arthralgia and/or myalgia in 12%, day 1 granulocytopenia in 3%, febrile neutropenia in 3%, and thromboembolic events in 1%. An additional 1% of PTX-treated patients experienced grade 3 or higher thromboembolic events subsequent to the completion of their chemotherapy. Severe hypersensitivity reactions occurred in 1% of patients during PTX administration. Incidence of grade 3 or higher cardiac dysfunction either during or subsequent to therapy was 1.0% in the control arm and 0.9% in the experimental arm. There were eight cases of acute myelogenous leukemia or myelodysplastic syndrome (AML/MDS). Six of the eight cases occurred in the AC PTX arm, and two in the AC arm. The interval from first chemotherapy to the development of the AML/MDS ranged from 9 months to 46 months. Five of the eight patients had received colony-stimulating factor, and five of the eight had also received radiotherapy as part of their treatment regimen.
DISCUSSION
Our results provide important confirmatory information regarding the benefit of incorporating a taxane (paclitaxel) in the adjuvant setting. This is the third large adjuvant trial reported to date comparing a taxane-containing with a non–taxane-containing regimen in the adjuvant setting. The other two reported trials include CALGB 9344, which had a similar design to our study and compared sequential paclitaxel after AC with AC alone, and Breast Cancer International Research Group (BCIRG) 001, which compared the TAC regimen (docetaxel, doxorubicin, and cyclophosphamide) to the FAC regimen (fluorouracil, doxorubicin, and cyclophosphamide). Mature results from CALGB 9344 with 69 months of median follow-up demonstrated that the addition of paclitaxel to AC significantly improved recurrence-free survival and OS (reduction in the hazard rate of recurrence with paclitaxel, 17%; P = .0023; reduction in the hazard rate of death, 18%; P = .0064). Although a possible interaction between paclitaxel effect and hormone-receptor status was reported with early follow-up in that study, no significant interaction was ultimately observed (after controlling for multiple comparisons) when the mature results were published. BCIRG 001 has been reported only in abstract form on the basis of results of the first24 and second interim analysis.25 On the basis of the second interim analysis, with 55 months of median follow-up, a significant benefit in DFS and OS was shown in favor of the TAC regimen compared with the FAC regimen. There was a 28% reduction in DFS events (5-year DFS, 75% with TAC and 68% with FAC; P = .001) and a 30% reduction in mortality (5-year OS, 87% with TAC and 81% with FAC; P = .008). No interaction between treatment effect and hormone-receptor status was observed in that trial as well.
Despite a significant improvement in DFS, no significant improvement in OS has been observed in our study with more than 64 months of median follow-up. There are several possible explanations for this observation. First and foremost, our results of a 7% reduction in death rate with the addition of paclitaxel are not inconsistent with a modest benefit in mortality with the addition of paclitaxel and, as importantly, are not statistically different from the mortality reductions observed in CALGB 9344. It is not uncommon for a significant difference in OS to emerge 2 to 3 years after a significant DFS difference is observed in an adjuvant trial. Furthermore, in our study the majority of the patients (70%) presented with one to three positive nodes, and only 4% presented with 10 or more positive nodes, making for a "better prognosis" cohort of node-positive patients than in CALGB 9344 (46% of patients with one to three positive nodes and 12% of patients with 10 positive nodes) or in BCIRG 001 (62% of patients with one to three positive nodes and 10% with 10 positive nodes). In fact, the 5-year OS for the AC arm in our study was 85% as opposed to 77% in CALGB 9344% and 81% in BCIRG 001 (FAC arm). Furthermore, the introduction of several important new drugs for advanced breast cancer (such as third-generation aromatase inhibitors, capecitabine, and trastuzumab) have significantly prolonged the survival of patients with advanced breast cancer,26 thus requiring more follow-up before differences in OS are realized in the adjuvant setting.
The sequential use of paclitaxel after AC led to a noticeable reduction in the incidence of contralateral breast cancer (P = .039; Table 4). Of interest, this putative effect was evidenced only in patients whose primary breast tumors were hormone-receptor-positive (data not shown). Whether these observations are the result of chance or are actually of substance may only be addressed by continued follow-up of this and other similar trials investigating the effect of taxanes in the treatment of early-stage breast cancer.
The dose of paclitaxel selected for our study (225 mg/m2) is higher than the approved paclitaxel dose for advanced breast cancer and the dose used in CALGB 9344 (175 mg/m2). This dose was selected based on available information at the time of inception of the B-28 trial suggesting that a dose-response relationship existed with paclitaxel in ovarian and breast cancer,18,19 and that the maximum-tolerated dose of paclitaxel as a 3-hour infusion was between 250 and 300 mg/m2. In fact, existence of a dose-response relationship was further suggested (but not definitively proven) in 1998 in another CALGB study (CALGB 9342) comparing three different doses of paclitaxel (175, 210 and 250 mg/m2) in patients with advanced breast cancer.20 That study showed a borderline statistically significant improvement in time to progression but no significant differences in response rates or OS with the higher doses of paclitaxel. However, when the results from our study are indirectly compared with those from CALGB 9344, the higher dose used in B-28 did not appear to be more effective than the lower dose used in CALGB 9344. One possible explanation for this observation may be that cumulative neurotoxicity increases with higher paclitaxel dose, leading to dose reductions and, in some patients, treatment discontinuation. Supportive of that hypothesis is the fact that fewer patients in B-28 completed all four cycles of paclitaxel (76%) when compared with CALGB 9344 (92%).7
In our study, tamoxifen was administered concurrently with chemotherapy. At the time of the design and conduct of the trial, no available information from randomized trials existed on the optimal administration of chemotherapy and hormonal therapy (ie, concurrent v sequential). At the time, there were theoretical arguments and preclinical data supporting the adoption of either approach.27-34 Recent results from a large Southwest Oncology Group (SWOG) trial demonstrated significant improvement in 8-year DFS when tamoxifen was administered after chemotherapy rather than concurrently.35 It is unlikely, however, that the concurrent administration of chemotherapy and hormonal therapy adversely influenced the benefit from paclitaxel in our study. First, the significant difference in DFS in the SWOG trial in favor of the sequential administration did not emerge until 8 years of follow-up. Second, a significant 23% reduction in DFS events was observed in our trial in the HR-positive subset, arguing against a negative interaction from the concurrent administration of the two modalities in that group of patients. Although this reduction was due in part to some imbalance in contralateral breast cancers and other second primary cancers, even the reduction in recurrence (12% in HR-positive patients in our study), was similar to the 9% reduction observed in the HR-positive cohort in CALGB 9344. Finally, indirect information can be obtained from a neoadjuvant trial (NSABP B-27) that compared AC with AC followed by docetaxel with concurrent administration of tamoxifen. In that trial, the addition of sequential docetaxel to AC almost doubled the rate of pathologic complete response (pCR) when compared with AC alone. Although one could argue that the difference in pCR could have been greater if tamoxifen was administered after completion of neoadjuvant chemotherapy, the rates of pCR with AC followed by docetaxel observed in B-27 and the differences in pCR between the taxane-containing and the non–taxane-containing arms are among the highest observed to date.36-41
Although in our study, as well as in CALGB 9344, there is no evidence of significant interaction between treatment effect and hormone-receptor status, the reduction in recurrence rate was of lesser magnitude in receptor-positive patients (12% and 9%, respectively) than in receptor-negative patients (17% and 28%, respectively). This observation is consistent with the results of the overview analysis demonstrating greater reductions in recurrence and mortality in the receptor-negative cohort than in the receptor-positive cohort.1 Recent results from several neoadjuvant trials also confirm that patients with receptor-negative tumors achieve higher pathologic response rates compared with patients with receptor-positive tumors.42,43 Thus, it appears that the relatively higher chemosensitivity of receptor-negative patients compared with receptor-positive patients is observed with many chemotherapy regimens and is not taxane specific. Interestingly, in the second interim analysis of the BCIRG 001 trial, similar, statistically significant reductions in recurrence with TAC compared with FAC were observed in both receptor-negative and in receptor-positive patients (31% v 28% respectively).25
Some have argued that the benefits observed in studies where taxanes are added in sequence to an anthracycline-based regimen might be due (at least in part) to the administration of more cycles of chemotherapy in the experimental arm (for example 8 in the AC PTX arm v 4 in the AC arm of B-28 and CALGB 9344). However, several additional adjuvant and neoadjuvant trials in which chemotherapy duration was the same in the control and in the experimental arm support the hypothesis that the additional benefit observed is the result of the administration of a non–cross-resistant agent rather than the result of additional chemotherapy cycles in the experimental group.24,44,45
On the basis of the results from the CALGB 9344 and the BCIRG 001 trials, the AC PTX regimen with paclitaxel administered at 175 mg/m2 and the TAC regimen were approved in the United States as adjuvant chemotherapy for node-positive breast cancer. Several second-generation trials have been designed to compare different anthracycline-taxane regimens as adjuvant therapy. Attempts to optimize the sequential anthracycline/paclitaxel regimen by administering it in a dose-dense fashion (every 2 weeks with colony-stimulating factor support) have produced early promising results in terms of further improving DFS and OS.46 In another attempt to optimize the efficacy and safety of the sequential anthracycline/paclitaxel regimen, neoadjuvant paclitaxel administered weekly followed by neoadjuvant FAC was compared with neoadjuvant paclitaxel administered in the traditional every-3-weeks fashion (as a 24-hour infusion) followed by the same neoadjuvant FAC. The weekly paclitaxel regimen resulted in a significant increase in the pCR rates (29% v 14%; P < .01) and improved safety profile.38 In a further attempt to identify the most optimal sequential anthracycline-taxane regimen, Eastern Cooperative Oncology Group (ECOG) E1199 has compared AC followed by paclitaxel (administered every 3 weeks v weekly) with AC followed by docetaxel (administered every 3 weeks v weekly). Results from this study are pending. Similarly, the NSABP has recently compared the sequential AC followed by docetaxel regimen (for four cycles each) with four cycles of the combination of doxorubicin/docetaxel (AT) and with four cycles of the triple combination of doxorubicin/docetaxel/cyclophosphamide (TAC). The results of these and other recently completed or current trials will provide valuable information about the safest and most effective way to incorporate taxanes in the adjuvant setting.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Consultant/Advisory Role: Eleftherios P. Mamounas, Aventis, Bristol-Myers Squibb. Honoraria: Eleftherios P. Mamounas, Aventis, Bristol-Myers Squibb; D. Lawrence Wickerham, AstraZeneca. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and Disclosures of Potential Conflicts of Interest found in Information for Contributors in the front of each issue.
Acknowledgment
We thank Barbara C. Good, PhD, for editorial assistance.
NOTES
Supported by Public Health Service grants U10CA-12027, U10CA-69974, U10CA-37377, and U10CA-69651 from the National Cancer Institute, Department of Health and Human Services, National Institutes of Health (NIH), Bethesda, MD.
Presented in abstract form at the 39th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 3, 2003; interim results were presented at the 2000 NIH Consensus Development Conference, Bethesda, MD, November 1-3, 2000.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Early Breast Cancer Trialists' Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trials. Early Breast Cancer Trialists' Collaborative Group. Lancet 352:930-942, 1998
Hortobagyi GN: Progress in systemic chemotherapy of primary breast cancer: An overview. J Natl Cancer Inst Monogr 30:72-79, 2001
Fisher B, Brown AM, Dimitrov NV, et al: Two months of doxorubicin-cyclophosphamide with and without interval reinduction therapy compared with 6 months of cyclophosphamide, methotrexate, and fluorouracil in positive-node breast cancer patients with tamoxifen-nonresponsive tumors: Results from the National Surgical Adjuvant Breast and Bowel Project B-15. J Clin Oncol 8:1483-1496, 1990
Fisher B, Redmond C, Legault-Poisson S, et al: Postoperative chemotherapy and tamoxifen compared with tamoxifen alone in the treatment of positive-node breast cancer patients aged 50 years and older with tumors responsive to tamoxifen: Results from the National Surgical Adjuvant Breast and Bowel Project B-16. J Clin Oncol 8:1005-1018, 1990
Fisher B, Anderson S, Wickerham DL, et al: Increased intensification and total dose of cyclophosphamide in a doxorubicin-cyclophosphamide regimen for the treatment of primary breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-22. J Clin Oncol 15:1858-1869, 1997
Fisher B, Anderson S, DeCillis A, et al: Further evaluation of intensified and increased total dose of cyclophosphamide for the treatment of primary breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-25. J Clin Oncol 17:3374-3388, 1999
Henderson IC, Berry DA, Demetri GD, et al: Improved outcomes from adding sequential paclitaxel but not from escalating doxorubicin dose in an adjuvant chemotherapy regimen for patients with node-positive primary breast cancer. J Clin Oncol 21:976-983, 2003
Holmes FA, Walters RS, Theriault RL, et al: Phase II trial of Taxol, an active drug in the treatment of metastatic breast cancer. J Natl Cancer Inst 83:1797-1805, 1991
Seidman AD, Tiersten A, Hudis C, et al: Phase II trial of paclitaxel by 3-hour infusion as initial and salvage chemotherapy for metastatic breast cancer. J Clin Oncol 13:2575-2581, 1995
Seidman AD, Reichman BS, Crown JP, et al: Paclitaxel as second and subsequent therapy for metastatic breast cancer: Activity independent of prior anthracycline response. J Clin Oncol 13:1152-1159, 1995
Wilson WH, Berg SL, Bryant G, et al: Paclitaxel in doxorubicin-refractory or mitoxantrone-refractory breast cancer: A phase I/II trial of 96-hour infusion. J Clin Oncol 12:1621-1629, 1994
Valero V, Holmes FA, Walters RS, et al: Phase II trial of docetaxel: A new, highly effective antineoplastic agent in the management of patients with anthracycline-resistant metastatic breast cancer. J Clin Oncol 13:2886-2894, 1995
Gianni L, Munzone E, Capri G, et al: Paclitaxel in metastatic breast cancer: A trial of two doses by a 3-hour infusion in patients with disease recurrence after prior therapy with anthracyclines. J Natl Cancer Inst 87:1169-1175, 1995
Vermorken JB, ten Bokkel Huinink WW, Mandjes IA, et al: High-dose paclitaxel with granulocyte colony-stimulating factor in patients with advanced breast cancer refractory to anthracycline therapy: A European Cancer Center trial. Semin Oncol 22:16-22, 1995
Schiller JH, Storer B, Tutsch K, et al: Phase I trial of 3-hour infusion of paclitaxel with or without granulocyte colony-stimulating factor in patients with advanced cancer. J Clin Oncol 12:241-248, 1994
Smith RE, Brown AM, Mamounas EP, et al: Randomized trial of 3-hour versus 24-hour infusion of high-dose paclitaxel in patients with metastatic or locally advanced breast cancer: National Surgical Adjuvant Breast and Bowel Project Protocol B-26. J Clin Oncol 17:3403-3411, 1999
Seidman AD: The emerging role of paclitaxel in breast cancer therapy. Clin Cancer Res 1:247-256, 1995
Sarosy G, Reed E: Taxol dose intensification and its clinical implications. J Natl Med Assoc 85:427-431, 1993
Nabholtz JM, Gelmon K, Bontenbal M, et al: Multicenter, randomized comparative study of two doses of paclitaxel in patients with metastatic breast cancer. J Clin Oncol 14:1858-1867, 1996
Winer E, Berry D, Henderson IC, et al: Failure of higher dose of paclitaxel to improve outcome in patients with metastatic breast cancer-Results from CALGB 9342. Proc Am Soc Clin Oncol 17:101A, 1998 (abstr 388)
White SJ, Freedman LS: Allocation of patients to treatment groups in a controlled clinical study. Br J Cancer 37:849-857, 1978
Fleming TR, Harrington DP, O'Brien PC: Designs for group sequential tests. Control Clin Trials 5:348-361, 1984
Korn EL, Dorey FJ: Applications of crude incidence curves. Stat Med 11:813-829, 1992
Nabholtz J-M, Pienkowski T, Mackey J, et al: Phase III trial comparing TAC (docetaxel, doxorubincin, cyclophosphamide) with FAC (5-fluorouracil, doxorubicin, cyclophosphamide) in the adjuvant treatment of node positive breast cancer (BC) patients: Interim analysis of the BCIRG 0001 study. Proc Am Soc Clin Oncol 21:36A, 2002 (abstr 141)
Martin M, Pienkowski T, Mackey J, et al: TAC improves disease free survival and overall survival over FAC in node positive early breast cancer patients, BCIRG 001: 55 months follow-up. Breast Cancer Res Treat 82, 2003 (abstr 43)
Chia SKL, Speers C, Kang A, et al: The impact of new chemotherapeutic and hormonal agents on the survival of women with metastatic breast cancer (MBC) in a population based cohort. Proc Am Soc Clin Oncol 22:6, 2003 (abstr 22)
Hug V, Hortobagyi GN, Drewinko B, et al: Tamoxifen-citrate counteracts the antitumor effects of cytotoxic drugs in vitro. J Clin Oncol 3:1672-1677, 1985
Foster BJ, Grotzinger KR, McKoy WM, et al: Modulation of induced resistance to adriamycin in two human breast cancer cell lines with tamoxifen or perhexiline maleate. Cancer Chemother Pharmacol 22:147-152, 1988
Osborne CK, Kitten L, Arteaga CL: Antagonism of chemotherapy-induced cytotoxicity for human breast cancer cells by antiestrogens. J Clin Oncol 7:710-717, 1989
Wiebe V, Koester S, Lindberg M, et al: Toremifene and its metabolites enhance doxorubicin accumulation in estrogen receptor negative multidrug resistant human breast cancer cells. Invest New Drugs 10:63-71, 1992
Woods KE, Randolph JK, Gewirtz DA: Antagonism between tamoxifen and doxorubicin in the MCF-7 human breast tumor cell line. Biochem Pharmacol 47:1449-1452, 1994
Leonessa F, Jacobson M, Boyle B, et al: Effect of tamoxifen on the multidrug-resistant phenotype in human breast cancer cells: Isobologram, drug accumulation, and M(r) 170,000 glycoprotein (gp170) binding studies. Cancer Res 54:441-447, 1994
Claudio JA, Emerman JT: The effects of cyclosporin A, tamoxifen, and medroxyprogesterone acetate on the enhancement of adriamycin cytotoxicity in primary cultures of human breast epithelial cells. Breast Cancer Res Treat 41:111-122, 1996
de Vincenzo R, Scambia G, Benedetti Panici P, et al: Chemosensitizing effect of tamoxifen and ICI 182,780 on parental and adriamycin-resistant MCF-7 human breast cancer cells. Ann N Y Acad Sci 784:517-520, 1996
Albain KS, Green SJ, Ravdin PM, et al: Adjuvant chemohormonal therapy for primary breast cancer should be sequential instead of concurrent: Initial results from Intergroup trial 0100 (SWOG 8814). Proc Am Soc Clin Oncol 21:37A, 2002 (abstr 143)
Gianni L, Baselga J, Eiermann W, et al: First report of the European Cooperative Trial in operable breast cancer (ECTO): Effects of primary systemic therapy (PST) on local-regional disease. Proc Am Soc Clin Oncol 21:34A, 2002 (abstr 132)
Smith IE, Jones AL, O'Brien ME, et al: Primary medical (neo-adjuvant) chemotherapy for operable breast cancer. Eur J Cancer 29A:1796-1799, 1993
Green M, Buzdar A, Smith T, et al: Weekly (wkly) paclitaxel (P) followed by FAC as primary systemic chemotherapy (PSC) of operable breast cancer improves pathologic complete remission (pCR) rates when compared to every 3-week (Q3 wk) P therapy (tx) followed by FAC-final results of a prospective phase III randomized trial. Proc Am Soc Clin Oncol 21:35A, 2002 (abstr 135)
Untch M, Kochecny G, Ditsch N, et al: Dose-dense sequential epirubicin-paclitaxel as preoperative treatment of breast cancer: Results of a randomized AGO study. Proc Am Soc Clin Oncol 21:34A, 2002 (abstr 133)
von Minckwitz G, Raab M, Schuette N, et al: Dose-dense versus sequential adriamycin/docetaxel combination as preoperative chemotherapy (pCHT) in operable breast cancer (T2-3, N0-2, M0)—Primary endpoint analysis of the GEPARDUO-study. Proc Am Soc Clin Oncol 21:43A, 2002 (abstr 168)
Evans T, Gould A, Foster E, et al: Phase III randomized trial of adriamycin (A) and docetaxel (D) versus A and cyclophosphamide (C) as primary medical therapy (PMT) in women with breast cancer: An ACCOG study. Proc Am Soc Clin Oncol 21:35A, 2002 (abstr 136)
Bear H: The effect on primary tumor response of adding sequential Taxotere to Adriamycin and cyclophosphamide: Preliminary results from NSABP protocol B-27. Br Cancer Res Treat 69:210, 2001 (abstr 5)
Untch M, Kahlert S, Moebus V, et al: Negative steriod receptors are a good predictor for response to preoperative chemotherapy in breast cancer (BC)-results of a randomised trial. Proc Am Soc Clin Oncol 22:9, 2003, (abstr 35)
Smith IC, Heys SD, Hutcheon AW, et al: Neoadjuvant chemotherapy in breast cancer: Significantly enhanced response with docetaxel. J Clin Oncol 20:1456-1466, 2002
Buzdar AU, Singletary SE, Valero V, et al: Evaluation of paclitaxel in adjuvant chemotherapy for patients with operable breast cancer: Preliminary data of a prospective randomized trial. Clin Cancer Res 8:1073-1079, 2002
Citron M, Berry D, Cirrincione C, et al: Superiority of dose-dense (DD) over conventional scheduling (CS) and equivalence of sequential (SC) vs. combination adjuvant chemotherapy (CC) for node-positive breast cancer (CALGB 9741, INT C9741). Breast Cancer Res Treat 76:S32, 2002 (abstr 15)(Eleftherios P. Mamounas, )
The University of Pittsburgh Medical Center
Allegheny General Hospital, Pittsburgh, PA
Aultman Health Foundation, Canton, OH
Kaiser Permanente, Vallejo, CA
Rose Medical Center, Denver, CO
ABSTRACT
PURPOSE: The primary aim of National Surgical Adjuvant Breast and Bowel Project (NSABP) B-28 was to determine whether four cycles of adjuvant paclitaxel (PTX) after four cycles of adjuvant doxorubicin/cyclophosphamide (AC) will prolong disease-free survival (DFS) and overall survival (OS) compared with four cycles of AC alone in patients with resected operable breast cancer and histologically positive axillary nodes.
PATIENTS AND METHODS: Between August 1995 and May 1998, 3,060 patients were randomly assigned (AC, 1,529; AC followed by PTX [AC PTX], 1,531). Patients 50 years and those younger than 50 years with estrogen receptor (ER) or progesterone receptor (PR) -positive tumors also received tamoxifen for 5 years, starting with the first dose of AC. Postlumpectomy radiotherapy was mandated. Postmastectomy or regional radiotherapy was prohibited. Median follow-up is 64.6 months.
RESULTS: The addition of PTX to AC significantly reduced the hazard for DFS event by 17% (relative risk [RR], 0.83; 95% CI, 0.72 to 0.95; P = .006). Five-year DFS was 76% ± 2% for patients randomly assigned to AC PTX compared with 72% ± 2% for those randomly assigned to AC. Improvement in OS was small and not statistically significant (RR, 0.93; 95% CI, 0.78 to 1.12; P = .46). Five-year OS was 85% ± 2% for both groups. Subset analysis of the effect of paclitaxel according to hormone receptors or tamoxifen administration did not reveal statistically significant interaction (for DFS, P = .30 and P = .44, respectively). Toxicity with the AC PTX regimen was acceptable for the adjuvant setting.
CONCLUSION: The addition of PTX to AC resulted in significant improvement in DFS but no significant improvement in OS with acceptable toxicity. No significant interaction between treatment effect and receptor status or tamoxifen administration was observed.
INTRODUCTION
The value of adjuvant chemotherapy has been convincingly demonstrated in breast cancer patients with involved axillary nodes.1 Results from the most recent published update of the overview analysis indicate that administration of adjuvant chemotherapy significantly reduced the risk of recurrence by 23.5% and the risk of death by 15.3%. According to the same overview, the 10-year recurrence-free survival for node-positive patients treated with adjuvant chemotherapy was 47.6% (for patients younger than 50 years) and 43.6% (for those 50 to 69 years of age). The 10-year overall survival (OS) was 53.8% and 48.6%, respectively. Thus, despite significant progress in the treatment of node-positive breast cancer, there is still considerable room for improvement. In the overview analysis, the adjuvant chemotherapy most commonly used consisted of CMF (cyclophosphamide, methotrexate, and fluorouracil) or CMF-like regimens as well as anthracycline-containing regimens (administered for four or six cycles). When compared with patients treated with non–anthracycline-containing regimens, patients treated with anthracycline-containing regimens had a statistically significant reduction in recurrence rates (12%) and mortality rates (11%),1 although it has been argued that not all of the anthracycline-containing regimens included in the overview analysis are equally effective.2
The most commonly used anthracycline-containing adjuvant chemotherapy regimen in the United States consists of four cycles of doxorubicin plus cyclophosphamide (AC) administered every 21 days. This regimen became popular when results from a large adjuvant National Surgical Adjuvant Breast and Bowel Project (NSABP) study in node-positive patients who were deemed tamoxifen unresponsive (protocol B-15) showed its equivalence in efficacy and toxicity to the classic CMF regimen administered for six cycles (oral cyclophosphamide day 1 through 14 and methotrexate and fluorouracil on days 1 and 8).3 In addition, the AC regimen combined with tamoxifen was found to be superior to tamoxifen alone in another NSABP study conducted in patients who were deemed tamoxifen responsive (protocol B-16).4 Attempts to optimize the AC regimen by intensifying and/or increasing the dose of cyclophosphamide or that of doxorubicin did not result in improved outcome.5-7 One of the advantages of the AC regimen over classic CMF is that it can be given in much shorter time (84 days v 168 days) with fewer chemotherapy visits (four v 12). Thus, the AC regimen became attractive not only in the clinic but also in research studies as a control regimen to which one could sequentially add new non–cross-resistant drugs while keeping the total duration of adjuvant therapy to about 6 months.
In the early 1990s, the demonstration of significant antitumor activity with taxanes in patients with advanced breast cancer8-12 marked a new era in breast cancer chemotherapy and brought hope for the development of new, more active adjuvant chemotherapy regimens. Although the original clinical development of paclitaxel (PTX) started in the 1970s, initial progress was impeded by the occurrence of severe hypersensitivity reactions. In the late 1980s and early 1990s, hypersensitivity reactions were controlled with appropriate premedication and by administering the drug over longer periods of time (24 to 96 hours). Although longer infusions demonstrated significant clinical activity with good safety profile, they were impractical for the adjuvant setting. Thus, several studies explored and eventually confirmed the safety and efficacy of a 3-hour infusion at low or high doses,9,13-16 with 250 to 300 mg/m2 as a maximum-tolerated dose without colony-stimulating factor support.17 Of interest, a dose-response relationship with PTX had been originally postulated from data in ovarian cancer18 and was later also suggested in breast cancer studies,19,20 although this relationship was not as clear when intermediate and higher doses of PTX were compared with doses near the maximum-tolerated dose.20
In August of 1995, the NSABP initiated protocol B-28 in order to evaluate the worth of PTX as adjuvant chemotherapy for patients with resected operable breast cancer and histologically positive nodes. This randomized trial was designed to determine whether the addition of PTX (given at 225 mg/m2 as a 3-hour infusion) following four cycles of AC would prolong disease-free survival (DFS) and OS when compared with the administration of four cycles of AC alone (Fig 1). In this report, we present the first definitive analysis of this study.
PATIENTS AND METHODS
Patient Eligibility
Eligible patients were required to have resected operable adenocarcinoma confined to the breast and ipsilateral axilla on clinical examination (cT1-3, cN0-1, cM0) and to be randomly assigned within 63 days from initial cytologic or histologic diagnosis. Patients must have undergone either lumpectomy with free margins plus axillary node dissection or modified radical mastectomy, and the tumor had to be invasive adenocarcinoma with at least one positive axillary lymph node on pathologic examination. Determination of the estrogen receptor (ER) and progesterone receptor (PR) status of the primary tumor had to be performed before random assignment. Patients were required to have normal hematologic parameters and adequate hepatic and renal function as well as at least a 10-year life expectancy excluding their diagnosis of cancer. Eligible patients signed an approved consent form conforming to federal and institutional guidelines. Patients with a previous history of invasive breast cancer or ductal carcinoma-in-situ (in either breast) were ineligible, as were patients who had received any prior radiation, chemotherapy, immunotherapy, or hormonal therapy for their present breast cancer.
Protocol Therapy
Patients assigned to the control arm (AC) were to receive doxorubicin 60 mg/m2 by slow intravenous (IV) push followed by cyclophosphamide 600 mg/m2 IV infusion over 30 minutes to 2 hours every 21 days for four cycles. Patients assigned to AC followed by PTX [AC PTX] were to receive four cycles of AC as in the control arm, followed by four additional 21-day cycles of PTX 225 mg/m2 IV as a 3-hour infusion on day 1 of each cycle (AC PTX). Before each cycle of PTX, patients were required to receive premedication with dexamethasone 20 mg PO (12 and 6 hours before PTX infusion), diphenhydramine 50 mg IV (1 hour before PTX infusion) and cimetidine 300 mg IV or ranitidine 50 mg IV (1 hour before PTX infusion). All patients who were ER- or PR-positive, or who were 50 years of age or older at the time of surgery were to also receive tamoxifen 20 mg/day for 5 years, beginning on day 1 of the first AC cycle. Patients surgically treated with lumpectomy were required to receive whole-breast irradiation at a dose of 5,000 cGY in 25 fractions, with an optional external beam boost to the tumor site. Radiotherapy was administered following the last cycle of chemotherapy and after recovery from any toxicity. Postmastectomy local chest wall radiotherapy or regional nodal radiotherapy (irrespective of surgical procedure) was not permitted. Primary prophylaxis with colony-stimulating factors was not permitted, but secondary prophylaxis with recombinant human granulocyte colony-stimulating factor (rHu G-CSF; 5 mg/kg/day) was mandated following a cycle complicated by prolonged neutropenia, febrile neutropenia, or grade 3 to 4 infection. Antibiotic prophylaxis with ciprofloxacin was added to G-CSF prophylaxis following a cycle with grade 3 to 4 infection or after two cycles with febrile neutropenia. Dose reductions were required following a cycle with febrile neutropenia or grade 3 to 4 infection despite G-CSF and antibiotic prophylaxis. Dose reductions for PTX were required following a cycle with grade 3 neurotoxicity. PTX therapy was discontinued in case of grade 4 neurotoxicity, or in case of prolonged grade 2 or 3 neurotoxicity failing to return to grade 1 within 2 weeks. Similarly, PTX therapy was discontinued in case of severe hypersensitivity reaction or grade 3 to 4 cardiotoxicity.
Entry Studies and Follow-Up Requirements
Before assignment, patients were required to undergo history and physical examination (including gynecologic examination for patients with an intact uterus and/or ovaries), hematologic studies and chemistries, chest x-ray (within 3 months of random assignment), and ECG and bilateral mammogram (within 1 year from random assignment). Although a multi-gated pooled cardiac nuclear scan (MUGA) was not required, if performed it had to show a left ventricular ejection fraction of 45%. Patients were also required to have history and physical examination along with hematologic studies and chemistries on day 1 before each cycle of chemotherapy and every 6 months for the first 5 years. Gynecologic examination (when applicable), chest x-ray, and bilateral mammogram (unilateral for mastectomy patients) were required yearly for the first 5 years. Only physical examination, gynecologic examination (when applicable), and mammogram were required annually after 5 years.
Statistical Design and Methodology
Patient assignment to the two treatment arms was balanced with respect to histologic nodal status (one to three, four to nine, 10 positive nodes), assigned tamoxifen administration (no, yes), type of surgery (mastectomy, lumpectomy) and institution, using a biased-coin minimization algorithm.21 Participating sites faxed the required entry materials to the NSABP Biostatistical Center (Pittsburgh, PA), and random assignment was performed centrally. Patients' treatment assignments were then faxed back to the sites. The study was originally designed to have at least 80% power to detect a 25% reduction in mortality rate due to the sequential administration of PTX, using a .05-level two-sided statistical test for equality of mortality rates in the two study arms. When the trial was opened on August 1, 1995, the sample size was specified at 2,450 patients. After the review of early data indicating a lower percentage of patients completing all four cycles of PTX chemotherapy than was originally projected, the target sample size was increased to 3,050 patients. Definitive analysis was scheduled to take place after the report of the 490th death on both treatment arms combined. This ensures a power of 80% to detect a 22.6% reduction in mortality rate, in order to provide for a potentially attenuated treatment effect caused by noncompletion of therapy. Interim efficacy analyses were scheduled at protocol-specified intervals and took place after 64, 120, 269, and 369 deaths were reported. Early stopping boundaries were specified by the method of Fleming, Harrington, and O'Brien.22 In order to account for early interim looks, the final analysis is based on a .0452 two-sided level of significance. Following review of the third interim analysis, the Independent Data Monitoring Committee (DMC), which oversees all NSABP treatment trials, recommended that the interim results be reviewed at the National Institutes of Health Consensus Development Conference on Adjuvant Therapy for Breast Cancer (Bethesda, MD, November 1-3, 2000) even though the formal interim analysis boundaries had not been crossed. The DMC clearly expressed that this analysis should not be considered definitive. With this exception, no efficacy data were presented outside the DMC until the end point of 490 reported deaths was achieved.
Per protocol, the primary end points for analysis were DFS and OS. Events used in the determination of DFS include local, regional, and distant treatment failures, contralateral breast cancers, nonbreast second primary cancers, and deaths before recurrence or second primary cancer. The OS end point is based on deaths from any cause. Treatment comparisons of these two end points were based on log-rank tests, stratified by nodal status (one to three, four to nine, 10 positive nodes), protocol-assigned tamoxifen administration (no, yes) and operation (lumpectomy, mastectomy). Relative risks (RRs) and 95% CIs were estimated by fitting Cox models, with a dichotomous covariate representing assigned treatment arm and stratified in the same way. The analyses reported here are based on the intent-to-treat principle, and thus include all patients with follow-up, whether eligible or not, and regardless of the treatment actually received. Secondary analyses based on the eligible cohort of patients with follow-up gave substantively identical results. As an additional diagnostic check, treatments were compared after controlling for additional prognostic variables (number of positive nodes, clinical tumor size, operation, hormone receptor status, age, tamoxifen treatment, tumor grade, histologic tumor type) by fitting Cox models.
Based on post hoc findings reported in the Cancer and Leukemia Group B (CALGB) 9344 study,7 we also investigated whether the effectiveness of sequential AC PTX, measured on the basis of hazard ratios relative to treatment with AC alone, differed in hormone receptor (HR) -negative when compared with HR-positive patients. These tests were repeated for both the DFS and OS end points. Patients were considered to be HR-positive if they were either ER positive or PR positive. Hazard ratios were estimated separately for HR-negative and HR-positive cohorts by fitting Cox models after stratification for nodal status, tamoxifen, and operation. Log-hazard ratios were then compared using a Wald test to provide a test of PTX-by-HR interaction. A similar procedure was used to test for treatment-by-tamoxifen interaction. As an exploratory exercise, Cox models were also fitted to DFS and OS to screen for other interactions between treatment effect and prognostic factors (nodal status, clinical tumor size, operation, age at surgery, tumor grade, histologic tumor type). Analyses of recurrence-free interval (time to first recurrence censoring contralateral breast cancers, nonbreast second primary cancers, and deaths before diagnosis of cancer) were also performed. This end point was not specified in the protocol.
The individual end points defining first events (local, regional, distant recurrences, contralateral breast cancers, nonbreast second primary cancers, deaths before diagnosis of cancer) were analyzed by constructing a site-of-first-event table summarizing the frequencies of first events by treatment arm. For each end point, 5-year cumulative incidences were computed using the usual nonparametric estimator,23 and crude hazards were compared across treatment arms using the log-rank test.
Analyses reported here include all follow-up received at the NSABP Biostatistical Center as of the closure of the March 31, 2003, summary file, at which time 498 deaths and 863 DFS events had been reported. Median follow-up at that time was 64.6 months and was well balanced across treatment arms (Table 1). As of this report, vital status after 5 years is known for approximately 79% of all accrued patients. Only one patient has contributed no follow-up.
RESULTS
Accrual, Eligibility, and Patient Characteristics
The trial opened in August of 1995 and closed in May of 1998 after a accruing a total of 3,060 patients (1,529 patients were randomly assigned to AC and 1,531 to AC PXT). Thirteen patients in the AC arm and 11 patients in the AC PTX arm have been declared ineligible for various reasons (Table 1). Of 1,529 patients randomly assigned to AC, 1,528 (99.9%) had positive follow-up, compared with all of the 1,531 (100%) patients randomly assigned to AC PTX. Median follow-up is 64.8 months in the AC arm and 64.4 months in the AC PTX arm. Vital status at 5 years is known for 78.7% of patients in the AC arm and 79.9% of patients in the AC PTX arm.
Patient and tumor characteristics were distributed evenly between the two groups (Table 2). About half of the patients were younger than 50 years, and 19% were 60 years of age or older. Approximately 59% of the patients had tumors 2 cm in maximum diameter. Seventy percent of the patients in each arm had one to three positive nodes on pathologic examination, and only 4% of the patients had 10 positive nodes. Slightly less than half of the patients were treated with lumpectomy. About two thirds of the patients had ER-positive tumors, and approximately 61% to 62% had PR-positive tumors. As per protocol design, 85% of the patients were assigned to tamoxifen.
DFS and OS
There were 463 DFS events in the AC arm and 400 in the AC PTX arm (Table 3). Analysis of DFS demonstrated that the addition of PTX significantly reduced the risk of a DFS event by 17% (RR, 0.83; 95% CI, 0.72 to 0.95; P = .006). The 5-year DFS for patients in the AC arm was 72% ± 2% compared with 76% ± 2% for those in the AC PTX arm (Fig 2A).
There were 255 deaths in the AC arm and 243 in the AC PTX arm (Table 3). Analysis of OS demonstrated a non–statistically significant 7% reduction in death rate with the addition of PTX (RR, 0.93; 95% CI, 0.78 to 1.12; P = .46). The 5-year OS was 85% ± 2% for patients in both arms (Fig 2B). Although it appears that the curves are starting to separate after year 6, the number of patients at risk at the sixth and seventh year is relatively small, making any comparisons between the two arms at those time points inadvisable.
Table 4 shows the site of failure for all first events tabulated by treatment arm and summarizes crude hazard rates and 5-year cumulative incidence rates at each site. In addition to reducing recurrence rates, Table 4 suggests a possible effect of PTX in reducing the incidence of contralateral breast cancer (P = .039). There was also some reduction in the diagnosis of nonbreast second primary cancers in the AC PTX arm, although this effect was not statistically significant (P = .11). The cumulative incidence of all first events was 28.3% for the AC arm and 24.4% for the AC PTX arm.
Treatment effects for DFS and OS were also estimated by fitting multivariate Cox models that control for nodal status, clinical tumor size, operation, HR status, age, treatment with tamoxifen, histologic grade, and histologic tumor type. Estimates of treatment effect were almost identical to the ones obtained by stratified log-rank tests. For DFS, RR = 0.82 (95% CI, 0.72 to 0.94; P = .005; Table 5). For OS, RR = 0.92 (95% CI, 0.77 to 1.10; P = .34; data not shown). All of the prognostic factors were statistically significant in both the DFS and OS models, with the exception of histologic tumor type, which was only marginally significant for DFS (P = .09) and OS (P = .09). Prognostic factors were taken as reported and were not subjected to central review. There was no evidence of interaction between treatment effect and any of the factors, except for clinical tumor size. Treatment-by-tumor-size interaction was significant for OS (P = .02) but not for DFS (P = .08). However, the nature of the interaction suggested by the data appeared to be unlikely: PTX appeared relatively beneficial for either small tumors ( 2 cm) or large tumors (> 4 cm), but not for intermediate size tumors. This pattern of interaction appears to be spurious.
Effect of Treatment by HRs
Based on previous observation from CALGB 9344 relative to the effect of PTX according HR status, we undertook a similar subset analysis in this study. Analysis of the effect of PTX on DFS and OS according to HR status demonstrated no significant interaction between the effect of PTX in HR-positive compared with HR-negative patients (interaction P value .30 for DFS and .82 for OS; Table 3). The relative reduction in DFS events was 23% for HR-positive patients (RR, 0.77; 95% CI, 0.65 to 0.92; P = .004) and 10% for HR-negative patients (RR, 0.90; 95% CI, 0.72 to 1.12; P = .33; Table 3). The relative reduction in deaths was 6% for HR-positive patients (RR, 0.94; 95% CI, 0.74 to 1.21, P = .64) and 10% for HR-negative patients (RR, 0.90; 95% CI, 0.70 to 1.17, P = .44; Table 3).
We also performed a subset analysis of the effect of PTX on recurrence-free interval according to HR-status. Here again there was no evidence of an interaction relative to the effect of PTX in HR-positive compared with HR-negative patients. The relative reduction in recurrence was 12% for HR-positive patients (RR, 0.88; 95% CI, 0.72 to 1.07; P = .19) and 17% for the HR-negative patients (RR, 0.83; 95% CI, 0.65 to 1.05; P = .11) with an interaction P value of .69. Notice that although the interaction did not approach statistical significance, its sense was generally consistent with the data from CALGB 9344.
In addition to the above analyses, we examined the effect of PTX on DFS and OS according to tamoxifen administration (data not shown). In these subset analyses, there was no evidence of a significant treatment-by tamoxifen interaction.
Chemotherapy Completion Rate, Nonprotocol Therapy, and Toxicity
Chemotherapy completion rate was high during AC chemotherapy, with more than 98% of the patients completing all four cycles of AC. One hundred thirty-four patients of the 1,531 randomly assigned to AC PTX (8.8%) did not start PTX; 88% of those were due to patient withdrawal as opposed to physician withdrawal. The chemotherapy completion rate dropped during administration of PTX with 75.9% of the patients completing all eight cycles of protocol chemotherapy. For patients in the AC PTX arm in which treatment with PTX began but was discontinued before completion of eight cycles of therapy, physician withdrawal was the reason in 53% of the cases, with patient withdrawal accounting for the other 47%.
Toxicity from protocol therapy was acceptable for the adjuvant setting. There were seven deaths where treatment could not be excluded as a contributing factor. Five occurred in patients who received AC only (pulmonary embolism in one, congestive heart failure in two, sepsis in one, and seizure in one) and two in patients who received AC and PTX (coronary artery disease in one, pulmonary embolism in one). Most common grade 3 or greater toxicity during PTX therapy (based on the highest toxicity grade reported during PTX cycles) included neurosensory toxicity in 15% of patients, neuromotor toxicity in 7%, (grade 3 or higher neurosensory or neuromotor toxicity in 18%), arthralgia and/or myalgia in 12%, day 1 granulocytopenia in 3%, febrile neutropenia in 3%, and thromboembolic events in 1%. An additional 1% of PTX-treated patients experienced grade 3 or higher thromboembolic events subsequent to the completion of their chemotherapy. Severe hypersensitivity reactions occurred in 1% of patients during PTX administration. Incidence of grade 3 or higher cardiac dysfunction either during or subsequent to therapy was 1.0% in the control arm and 0.9% in the experimental arm. There were eight cases of acute myelogenous leukemia or myelodysplastic syndrome (AML/MDS). Six of the eight cases occurred in the AC PTX arm, and two in the AC arm. The interval from first chemotherapy to the development of the AML/MDS ranged from 9 months to 46 months. Five of the eight patients had received colony-stimulating factor, and five of the eight had also received radiotherapy as part of their treatment regimen.
DISCUSSION
Our results provide important confirmatory information regarding the benefit of incorporating a taxane (paclitaxel) in the adjuvant setting. This is the third large adjuvant trial reported to date comparing a taxane-containing with a non–taxane-containing regimen in the adjuvant setting. The other two reported trials include CALGB 9344, which had a similar design to our study and compared sequential paclitaxel after AC with AC alone, and Breast Cancer International Research Group (BCIRG) 001, which compared the TAC regimen (docetaxel, doxorubicin, and cyclophosphamide) to the FAC regimen (fluorouracil, doxorubicin, and cyclophosphamide). Mature results from CALGB 9344 with 69 months of median follow-up demonstrated that the addition of paclitaxel to AC significantly improved recurrence-free survival and OS (reduction in the hazard rate of recurrence with paclitaxel, 17%; P = .0023; reduction in the hazard rate of death, 18%; P = .0064). Although a possible interaction between paclitaxel effect and hormone-receptor status was reported with early follow-up in that study, no significant interaction was ultimately observed (after controlling for multiple comparisons) when the mature results were published. BCIRG 001 has been reported only in abstract form on the basis of results of the first24 and second interim analysis.25 On the basis of the second interim analysis, with 55 months of median follow-up, a significant benefit in DFS and OS was shown in favor of the TAC regimen compared with the FAC regimen. There was a 28% reduction in DFS events (5-year DFS, 75% with TAC and 68% with FAC; P = .001) and a 30% reduction in mortality (5-year OS, 87% with TAC and 81% with FAC; P = .008). No interaction between treatment effect and hormone-receptor status was observed in that trial as well.
Despite a significant improvement in DFS, no significant improvement in OS has been observed in our study with more than 64 months of median follow-up. There are several possible explanations for this observation. First and foremost, our results of a 7% reduction in death rate with the addition of paclitaxel are not inconsistent with a modest benefit in mortality with the addition of paclitaxel and, as importantly, are not statistically different from the mortality reductions observed in CALGB 9344. It is not uncommon for a significant difference in OS to emerge 2 to 3 years after a significant DFS difference is observed in an adjuvant trial. Furthermore, in our study the majority of the patients (70%) presented with one to three positive nodes, and only 4% presented with 10 or more positive nodes, making for a "better prognosis" cohort of node-positive patients than in CALGB 9344 (46% of patients with one to three positive nodes and 12% of patients with 10 positive nodes) or in BCIRG 001 (62% of patients with one to three positive nodes and 10% with 10 positive nodes). In fact, the 5-year OS for the AC arm in our study was 85% as opposed to 77% in CALGB 9344% and 81% in BCIRG 001 (FAC arm). Furthermore, the introduction of several important new drugs for advanced breast cancer (such as third-generation aromatase inhibitors, capecitabine, and trastuzumab) have significantly prolonged the survival of patients with advanced breast cancer,26 thus requiring more follow-up before differences in OS are realized in the adjuvant setting.
The sequential use of paclitaxel after AC led to a noticeable reduction in the incidence of contralateral breast cancer (P = .039; Table 4). Of interest, this putative effect was evidenced only in patients whose primary breast tumors were hormone-receptor-positive (data not shown). Whether these observations are the result of chance or are actually of substance may only be addressed by continued follow-up of this and other similar trials investigating the effect of taxanes in the treatment of early-stage breast cancer.
The dose of paclitaxel selected for our study (225 mg/m2) is higher than the approved paclitaxel dose for advanced breast cancer and the dose used in CALGB 9344 (175 mg/m2). This dose was selected based on available information at the time of inception of the B-28 trial suggesting that a dose-response relationship existed with paclitaxel in ovarian and breast cancer,18,19 and that the maximum-tolerated dose of paclitaxel as a 3-hour infusion was between 250 and 300 mg/m2. In fact, existence of a dose-response relationship was further suggested (but not definitively proven) in 1998 in another CALGB study (CALGB 9342) comparing three different doses of paclitaxel (175, 210 and 250 mg/m2) in patients with advanced breast cancer.20 That study showed a borderline statistically significant improvement in time to progression but no significant differences in response rates or OS with the higher doses of paclitaxel. However, when the results from our study are indirectly compared with those from CALGB 9344, the higher dose used in B-28 did not appear to be more effective than the lower dose used in CALGB 9344. One possible explanation for this observation may be that cumulative neurotoxicity increases with higher paclitaxel dose, leading to dose reductions and, in some patients, treatment discontinuation. Supportive of that hypothesis is the fact that fewer patients in B-28 completed all four cycles of paclitaxel (76%) when compared with CALGB 9344 (92%).7
In our study, tamoxifen was administered concurrently with chemotherapy. At the time of the design and conduct of the trial, no available information from randomized trials existed on the optimal administration of chemotherapy and hormonal therapy (ie, concurrent v sequential). At the time, there were theoretical arguments and preclinical data supporting the adoption of either approach.27-34 Recent results from a large Southwest Oncology Group (SWOG) trial demonstrated significant improvement in 8-year DFS when tamoxifen was administered after chemotherapy rather than concurrently.35 It is unlikely, however, that the concurrent administration of chemotherapy and hormonal therapy adversely influenced the benefit from paclitaxel in our study. First, the significant difference in DFS in the SWOG trial in favor of the sequential administration did not emerge until 8 years of follow-up. Second, a significant 23% reduction in DFS events was observed in our trial in the HR-positive subset, arguing against a negative interaction from the concurrent administration of the two modalities in that group of patients. Although this reduction was due in part to some imbalance in contralateral breast cancers and other second primary cancers, even the reduction in recurrence (12% in HR-positive patients in our study), was similar to the 9% reduction observed in the HR-positive cohort in CALGB 9344. Finally, indirect information can be obtained from a neoadjuvant trial (NSABP B-27) that compared AC with AC followed by docetaxel with concurrent administration of tamoxifen. In that trial, the addition of sequential docetaxel to AC almost doubled the rate of pathologic complete response (pCR) when compared with AC alone. Although one could argue that the difference in pCR could have been greater if tamoxifen was administered after completion of neoadjuvant chemotherapy, the rates of pCR with AC followed by docetaxel observed in B-27 and the differences in pCR between the taxane-containing and the non–taxane-containing arms are among the highest observed to date.36-41
Although in our study, as well as in CALGB 9344, there is no evidence of significant interaction between treatment effect and hormone-receptor status, the reduction in recurrence rate was of lesser magnitude in receptor-positive patients (12% and 9%, respectively) than in receptor-negative patients (17% and 28%, respectively). This observation is consistent with the results of the overview analysis demonstrating greater reductions in recurrence and mortality in the receptor-negative cohort than in the receptor-positive cohort.1 Recent results from several neoadjuvant trials also confirm that patients with receptor-negative tumors achieve higher pathologic response rates compared with patients with receptor-positive tumors.42,43 Thus, it appears that the relatively higher chemosensitivity of receptor-negative patients compared with receptor-positive patients is observed with many chemotherapy regimens and is not taxane specific. Interestingly, in the second interim analysis of the BCIRG 001 trial, similar, statistically significant reductions in recurrence with TAC compared with FAC were observed in both receptor-negative and in receptor-positive patients (31% v 28% respectively).25
Some have argued that the benefits observed in studies where taxanes are added in sequence to an anthracycline-based regimen might be due (at least in part) to the administration of more cycles of chemotherapy in the experimental arm (for example 8 in the AC PTX arm v 4 in the AC arm of B-28 and CALGB 9344). However, several additional adjuvant and neoadjuvant trials in which chemotherapy duration was the same in the control and in the experimental arm support the hypothesis that the additional benefit observed is the result of the administration of a non–cross-resistant agent rather than the result of additional chemotherapy cycles in the experimental group.24,44,45
On the basis of the results from the CALGB 9344 and the BCIRG 001 trials, the AC PTX regimen with paclitaxel administered at 175 mg/m2 and the TAC regimen were approved in the United States as adjuvant chemotherapy for node-positive breast cancer. Several second-generation trials have been designed to compare different anthracycline-taxane regimens as adjuvant therapy. Attempts to optimize the sequential anthracycline/paclitaxel regimen by administering it in a dose-dense fashion (every 2 weeks with colony-stimulating factor support) have produced early promising results in terms of further improving DFS and OS.46 In another attempt to optimize the efficacy and safety of the sequential anthracycline/paclitaxel regimen, neoadjuvant paclitaxel administered weekly followed by neoadjuvant FAC was compared with neoadjuvant paclitaxel administered in the traditional every-3-weeks fashion (as a 24-hour infusion) followed by the same neoadjuvant FAC. The weekly paclitaxel regimen resulted in a significant increase in the pCR rates (29% v 14%; P < .01) and improved safety profile.38 In a further attempt to identify the most optimal sequential anthracycline-taxane regimen, Eastern Cooperative Oncology Group (ECOG) E1199 has compared AC followed by paclitaxel (administered every 3 weeks v weekly) with AC followed by docetaxel (administered every 3 weeks v weekly). Results from this study are pending. Similarly, the NSABP has recently compared the sequential AC followed by docetaxel regimen (for four cycles each) with four cycles of the combination of doxorubicin/docetaxel (AT) and with four cycles of the triple combination of doxorubicin/docetaxel/cyclophosphamide (TAC). The results of these and other recently completed or current trials will provide valuable information about the safest and most effective way to incorporate taxanes in the adjuvant setting.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Consultant/Advisory Role: Eleftherios P. Mamounas, Aventis, Bristol-Myers Squibb. Honoraria: Eleftherios P. Mamounas, Aventis, Bristol-Myers Squibb; D. Lawrence Wickerham, AstraZeneca. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and Disclosures of Potential Conflicts of Interest found in Information for Contributors in the front of each issue.
Acknowledgment
We thank Barbara C. Good, PhD, for editorial assistance.
NOTES
Supported by Public Health Service grants U10CA-12027, U10CA-69974, U10CA-37377, and U10CA-69651 from the National Cancer Institute, Department of Health and Human Services, National Institutes of Health (NIH), Bethesda, MD.
Presented in abstract form at the 39th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 3, 2003; interim results were presented at the 2000 NIH Consensus Development Conference, Bethesda, MD, November 1-3, 2000.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Early Breast Cancer Trialists' Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trials. Early Breast Cancer Trialists' Collaborative Group. Lancet 352:930-942, 1998
Hortobagyi GN: Progress in systemic chemotherapy of primary breast cancer: An overview. J Natl Cancer Inst Monogr 30:72-79, 2001
Fisher B, Brown AM, Dimitrov NV, et al: Two months of doxorubicin-cyclophosphamide with and without interval reinduction therapy compared with 6 months of cyclophosphamide, methotrexate, and fluorouracil in positive-node breast cancer patients with tamoxifen-nonresponsive tumors: Results from the National Surgical Adjuvant Breast and Bowel Project B-15. J Clin Oncol 8:1483-1496, 1990
Fisher B, Redmond C, Legault-Poisson S, et al: Postoperative chemotherapy and tamoxifen compared with tamoxifen alone in the treatment of positive-node breast cancer patients aged 50 years and older with tumors responsive to tamoxifen: Results from the National Surgical Adjuvant Breast and Bowel Project B-16. J Clin Oncol 8:1005-1018, 1990
Fisher B, Anderson S, Wickerham DL, et al: Increased intensification and total dose of cyclophosphamide in a doxorubicin-cyclophosphamide regimen for the treatment of primary breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-22. J Clin Oncol 15:1858-1869, 1997
Fisher B, Anderson S, DeCillis A, et al: Further evaluation of intensified and increased total dose of cyclophosphamide for the treatment of primary breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-25. J Clin Oncol 17:3374-3388, 1999
Henderson IC, Berry DA, Demetri GD, et al: Improved outcomes from adding sequential paclitaxel but not from escalating doxorubicin dose in an adjuvant chemotherapy regimen for patients with node-positive primary breast cancer. J Clin Oncol 21:976-983, 2003
Holmes FA, Walters RS, Theriault RL, et al: Phase II trial of Taxol, an active drug in the treatment of metastatic breast cancer. J Natl Cancer Inst 83:1797-1805, 1991
Seidman AD, Tiersten A, Hudis C, et al: Phase II trial of paclitaxel by 3-hour infusion as initial and salvage chemotherapy for metastatic breast cancer. J Clin Oncol 13:2575-2581, 1995
Seidman AD, Reichman BS, Crown JP, et al: Paclitaxel as second and subsequent therapy for metastatic breast cancer: Activity independent of prior anthracycline response. J Clin Oncol 13:1152-1159, 1995
Wilson WH, Berg SL, Bryant G, et al: Paclitaxel in doxorubicin-refractory or mitoxantrone-refractory breast cancer: A phase I/II trial of 96-hour infusion. J Clin Oncol 12:1621-1629, 1994
Valero V, Holmes FA, Walters RS, et al: Phase II trial of docetaxel: A new, highly effective antineoplastic agent in the management of patients with anthracycline-resistant metastatic breast cancer. J Clin Oncol 13:2886-2894, 1995
Gianni L, Munzone E, Capri G, et al: Paclitaxel in metastatic breast cancer: A trial of two doses by a 3-hour infusion in patients with disease recurrence after prior therapy with anthracyclines. J Natl Cancer Inst 87:1169-1175, 1995
Vermorken JB, ten Bokkel Huinink WW, Mandjes IA, et al: High-dose paclitaxel with granulocyte colony-stimulating factor in patients with advanced breast cancer refractory to anthracycline therapy: A European Cancer Center trial. Semin Oncol 22:16-22, 1995
Schiller JH, Storer B, Tutsch K, et al: Phase I trial of 3-hour infusion of paclitaxel with or without granulocyte colony-stimulating factor in patients with advanced cancer. J Clin Oncol 12:241-248, 1994
Smith RE, Brown AM, Mamounas EP, et al: Randomized trial of 3-hour versus 24-hour infusion of high-dose paclitaxel in patients with metastatic or locally advanced breast cancer: National Surgical Adjuvant Breast and Bowel Project Protocol B-26. J Clin Oncol 17:3403-3411, 1999
Seidman AD: The emerging role of paclitaxel in breast cancer therapy. Clin Cancer Res 1:247-256, 1995
Sarosy G, Reed E: Taxol dose intensification and its clinical implications. J Natl Med Assoc 85:427-431, 1993
Nabholtz JM, Gelmon K, Bontenbal M, et al: Multicenter, randomized comparative study of two doses of paclitaxel in patients with metastatic breast cancer. J Clin Oncol 14:1858-1867, 1996
Winer E, Berry D, Henderson IC, et al: Failure of higher dose of paclitaxel to improve outcome in patients with metastatic breast cancer-Results from CALGB 9342. Proc Am Soc Clin Oncol 17:101A, 1998 (abstr 388)
White SJ, Freedman LS: Allocation of patients to treatment groups in a controlled clinical study. Br J Cancer 37:849-857, 1978
Fleming TR, Harrington DP, O'Brien PC: Designs for group sequential tests. Control Clin Trials 5:348-361, 1984
Korn EL, Dorey FJ: Applications of crude incidence curves. Stat Med 11:813-829, 1992
Nabholtz J-M, Pienkowski T, Mackey J, et al: Phase III trial comparing TAC (docetaxel, doxorubincin, cyclophosphamide) with FAC (5-fluorouracil, doxorubicin, cyclophosphamide) in the adjuvant treatment of node positive breast cancer (BC) patients: Interim analysis of the BCIRG 0001 study. Proc Am Soc Clin Oncol 21:36A, 2002 (abstr 141)
Martin M, Pienkowski T, Mackey J, et al: TAC improves disease free survival and overall survival over FAC in node positive early breast cancer patients, BCIRG 001: 55 months follow-up. Breast Cancer Res Treat 82, 2003 (abstr 43)
Chia SKL, Speers C, Kang A, et al: The impact of new chemotherapeutic and hormonal agents on the survival of women with metastatic breast cancer (MBC) in a population based cohort. Proc Am Soc Clin Oncol 22:6, 2003 (abstr 22)
Hug V, Hortobagyi GN, Drewinko B, et al: Tamoxifen-citrate counteracts the antitumor effects of cytotoxic drugs in vitro. J Clin Oncol 3:1672-1677, 1985
Foster BJ, Grotzinger KR, McKoy WM, et al: Modulation of induced resistance to adriamycin in two human breast cancer cell lines with tamoxifen or perhexiline maleate. Cancer Chemother Pharmacol 22:147-152, 1988
Osborne CK, Kitten L, Arteaga CL: Antagonism of chemotherapy-induced cytotoxicity for human breast cancer cells by antiestrogens. J Clin Oncol 7:710-717, 1989
Wiebe V, Koester S, Lindberg M, et al: Toremifene and its metabolites enhance doxorubicin accumulation in estrogen receptor negative multidrug resistant human breast cancer cells. Invest New Drugs 10:63-71, 1992
Woods KE, Randolph JK, Gewirtz DA: Antagonism between tamoxifen and doxorubicin in the MCF-7 human breast tumor cell line. Biochem Pharmacol 47:1449-1452, 1994
Leonessa F, Jacobson M, Boyle B, et al: Effect of tamoxifen on the multidrug-resistant phenotype in human breast cancer cells: Isobologram, drug accumulation, and M(r) 170,000 glycoprotein (gp170) binding studies. Cancer Res 54:441-447, 1994
Claudio JA, Emerman JT: The effects of cyclosporin A, tamoxifen, and medroxyprogesterone acetate on the enhancement of adriamycin cytotoxicity in primary cultures of human breast epithelial cells. Breast Cancer Res Treat 41:111-122, 1996
de Vincenzo R, Scambia G, Benedetti Panici P, et al: Chemosensitizing effect of tamoxifen and ICI 182,780 on parental and adriamycin-resistant MCF-7 human breast cancer cells. Ann N Y Acad Sci 784:517-520, 1996
Albain KS, Green SJ, Ravdin PM, et al: Adjuvant chemohormonal therapy for primary breast cancer should be sequential instead of concurrent: Initial results from Intergroup trial 0100 (SWOG 8814). Proc Am Soc Clin Oncol 21:37A, 2002 (abstr 143)
Gianni L, Baselga J, Eiermann W, et al: First report of the European Cooperative Trial in operable breast cancer (ECTO): Effects of primary systemic therapy (PST) on local-regional disease. Proc Am Soc Clin Oncol 21:34A, 2002 (abstr 132)
Smith IE, Jones AL, O'Brien ME, et al: Primary medical (neo-adjuvant) chemotherapy for operable breast cancer. Eur J Cancer 29A:1796-1799, 1993
Green M, Buzdar A, Smith T, et al: Weekly (wkly) paclitaxel (P) followed by FAC as primary systemic chemotherapy (PSC) of operable breast cancer improves pathologic complete remission (pCR) rates when compared to every 3-week (Q3 wk) P therapy (tx) followed by FAC-final results of a prospective phase III randomized trial. Proc Am Soc Clin Oncol 21:35A, 2002 (abstr 135)
Untch M, Kochecny G, Ditsch N, et al: Dose-dense sequential epirubicin-paclitaxel as preoperative treatment of breast cancer: Results of a randomized AGO study. Proc Am Soc Clin Oncol 21:34A, 2002 (abstr 133)
von Minckwitz G, Raab M, Schuette N, et al: Dose-dense versus sequential adriamycin/docetaxel combination as preoperative chemotherapy (pCHT) in operable breast cancer (T2-3, N0-2, M0)—Primary endpoint analysis of the GEPARDUO-study. Proc Am Soc Clin Oncol 21:43A, 2002 (abstr 168)
Evans T, Gould A, Foster E, et al: Phase III randomized trial of adriamycin (A) and docetaxel (D) versus A and cyclophosphamide (C) as primary medical therapy (PMT) in women with breast cancer: An ACCOG study. Proc Am Soc Clin Oncol 21:35A, 2002 (abstr 136)
Bear H: The effect on primary tumor response of adding sequential Taxotere to Adriamycin and cyclophosphamide: Preliminary results from NSABP protocol B-27. Br Cancer Res Treat 69:210, 2001 (abstr 5)
Untch M, Kahlert S, Moebus V, et al: Negative steriod receptors are a good predictor for response to preoperative chemotherapy in breast cancer (BC)-results of a randomised trial. Proc Am Soc Clin Oncol 22:9, 2003, (abstr 35)
Smith IC, Heys SD, Hutcheon AW, et al: Neoadjuvant chemotherapy in breast cancer: Significantly enhanced response with docetaxel. J Clin Oncol 20:1456-1466, 2002
Buzdar AU, Singletary SE, Valero V, et al: Evaluation of paclitaxel in adjuvant chemotherapy for patients with operable breast cancer: Preliminary data of a prospective randomized trial. Clin Cancer Res 8:1073-1079, 2002
Citron M, Berry D, Cirrincione C, et al: Superiority of dose-dense (DD) over conventional scheduling (CS) and equivalence of sequential (SC) vs. combination adjuvant chemotherapy (CC) for node-positive breast cancer (CALGB 9741, INT C9741). Breast Cancer Res Treat 76:S32, 2002 (abstr 15)(Eleftherios P. Mamounas, )