A Phase III trial of sequential adjuvant chemotherapy for operable breast carcinoma

Final analysis with 10-year follow-up

Authors


Abstract

BACKGROUND

The current study was performed to assess whether sequential potentially noncross-resistant chemotherapy prolongs disease-free survival (DFS) and overall survival (OS) in patients with operable breast carcinoma.

METHODS

Seven hundred eighty-nine patients were registered and followed for a median of 10 years. They were treated in two groups. In Group 1, patients age < 50 years or age > 50 years but with either negative or unknown estrogen receptor (ER) status were randomized to receive 6 cycles of 5-fluorouracil, doxorubicin, and cyclophosphamide (FAC) alone or followed by 4 cycles of methotrexate and vinblastine (MV). In Group 2, patients age ≥ 50 years with ER-positive disease were randomized to receive either tamoxifen or combination chemotherapy (FAC + MV) for 10 cycles. Analysis was performed according to allocated treatment (intention to treat), with all randomized patients included.

RESULTS

In Group 1 there were no significant differences with regard to DFS or OS between the two treatment arms. The DFS at 5 years was 0.70 and 0.76, respectively, for FAC compared with FAC+MV (P = 0.26). The OS was similar for both groups (0.84 vs. 0.83). It is interesting to note that there was a statistically nonsignificant trend for improved DFS in the FAC + MV arm for patients who were ER-positive. In Group 2, tamoxifen alone led to more prolonged DFS compared to FAC+MV (0.78 vs. 0.66, respectively) but this did not reach statistical significance (P = 0.28). OS also was associated with a trend (P = 0.86) toward prolonged survival for the tamoxifen arm compared with the FAC+MV arm (0.85 vs. 0.74, respectively).

CONCLUSIONS

The results of the current trial concerning sequential adjuvant chemotherapy for operable breast carcinoma, which to our knowledge contains the longest follow-up presented to date, failed to demonstrate any additional benefit from the addition of 4 cycles of MV to 6 cycles of FAC chemotherapy. Cancer 2003;97:2716–23. © 2003 American Cancer Society.

DOI 10.1002/cncr.11396

It is current practice to recommend adjuvant systemic therapy for the vast majority of women who have undergone surgical treatment of operable breast carcinoma. Although patient selection remains heavily debated, it generally is accepted that women with hormone-responsive disease (estrogen receptor [ER]-positive/negative and progesterone receptor [PR]-positive) will potentially benefit from endocrine treatment. For those patients whose primary tumors do not express ER or PR, the only available treatment option is adjuvant chemotherapy.

Although the type and total duration of chemotherapy still is being explored in clinical trials, there is growing consensus regarding certain issues.1, 2 The Oxford overview analysis on polychemotherapy for early breast carcinoma, which compiles data from 47 trials involving 18,000 patients, reported a significant reduction in disease recurrence and overall mortality for both premenopausal and postmenopausal women.3 More interestingly, the clinical benefits were of similar magnitude for patients with both lymph node-positive and lymph node-negative disease. Anthracycline-containing regimens appeared to provide additional benefit (3% improvement in absolute 5-year mortality) compared with nonanthracycline regimens (primarily a variety of the cyclophosphamide, methotrexate, and 5-fluorouracil [CMF] schedule). With regard to the duration of adjuvant chemotherapy, the overview analysis demonstrated that > 6 months of the same polychemotherapy resulted in a nonsignificant 7% decrease in disease-free survival (DFS) but had no apparent impact on overall survival.

However, it remains unanswered whether sequential noncross-resistant chemotherapy confers any additional clinical benefit. In that regard, the addition of taxanes (paclitaxel and docetaxel) in sequence to anthracycline-containing regimens has been studied with promising, yet controversial, activity.4–6 The concept is not new. In the pretaxane era, sequential combinations of active chemotherapeutics with different antineoplastic mechanisms (and thus no or less cross-resistance and no overlapping toxicity) have been investigated in the metastatic as well as the adjuvant setting without much success.7, 8

At the time of inception of the current trial (1986), it was standard practice to recommend hormone therapy alone for postmenopausal patients with ER-positive/negative and PR-positive disease.9 The optimal duration of chemotherapy for ER-negative patients was unknown. Combination chemotherapy already had been found to confer a clinical advantage over single-agent therapy and the CMF regimen was prescribed most often. It is interesting to note that at the time, there was no conclusive evidence that combined chemohormonal therapy would add to hormonal therapy alone for hormone-responsive disease in postmenopausal women.

At the M. D. Anderson Cancer Center, we have been investigating adjuvant chemotherapy regimens based on doxorubicin since 1974.10–12 We conducted the current trial to evaluate the role of additional chemotherapy with vinblastine and methotrexate (which will be abbreviated as MV to be consistent with an earlier publication abbreviation used) on completion of six cycles of a chemotherapy regimen comprised of 5-fluorouracil, doxorubicin, and cyclophosphamide (FAC) in women with operable breast carcinoma who were either age < 50 years or age > 50 years but had ER-negative disease or unknown ER status for their primary tumor. The concept tested was that additional chemotherapy with different antineoplastic mechanisms and no overlapping toxicity or cross-resistance would be more beneficial than six cycles of doxorubicin-based therapy. In addition, we compared tamoxifen with sequential combination chemotherapy (FAC + MV) for women age > 50 years with ER-positive disease.

MATERIALS AND METHODS

Between 1986 and 1994, all patients age < 75 years with Stage II or resectable Stage III (AJCC staging system) breast carcinoma with or without lymph node involvement were eligible for the study if they had no evidence of metastatic disease at the time of initial evaluation. Patients with Stage I disease also were included if the tumors were ER-negative or expressed other histologic features of aggressive disease (lymphovascular invasion, high labeling index, or high nuclear grade). Patients treated with segmental mastectomy and irradiation also were eligible. Patients were excluded from the study if they had symptomatic cardiovascular disease, a history of other neoplasms (except squamous cell carcinoma of the skin or in situ cervical carcinoma), or a previous history of contralateral breast carcinoma with a higher stage of disease. Stage I patients with ER-positive primary tumors measuring < 1 cm (T1a, T1b, and N0 disease) and no poor prognostic histologic criteria (as stated earlier) also were excluded. This research was reviewed and approved by our institutional review board. Before enrollment, each patient was informed about the investigational nature of the study in keeping with the policy of the institution and written consent was obtained.

Pretreatment and Subsequent Evaluations

Before entry on the study, each patient's disease was staged by the following: a complete history and physical examination, compete blood count (CBC), serum chemistries, urinalysis, serum carcinoembryonic antigen, skeletal X -rays, chest X-ray, liver ultrasonography or nuclear scan, bone scan, and mammography of the contralateral breast. During therapy, the CBC was repeated weekly, and chemistries were obtained prior to each chemotherapy cycle. Diagnostic imaging studies were repeated every 3–4 months the first 2 years, biannually during the third year, and annually since unless clinically indicated. Mammograms were repeated annually. A cardiac evaluation was performed by noninvasive and/or invasive studies before the administration of further chemotherapy when there was clinical indication of cardiac dysfunction.

Treatment Plan

Each patient was registered prospectively with the data management office of the institution and all patients were randomized on Day 1 to one of three different treatment schemas depending on age and ER status. Patients age < 50 years or those who were age > 50 years but with either negative or unknown ER status were randomized to receive either FAC chemotherapy for 6 cycles or to the same chemotherapy regimen followed by 4 cycles of MV. Patients age > 50 years with ER-positive disease were randomized to receive either tamoxifen, 20 mg daily, or combination chemotherapy for 10 cycles (FAC ×6 followed by MV ×4). Before randomization, patients were stratified by menopausal status, stage of disease, number of involved lymph nodes, irradiation status, and hormone receptor status.

FAC chemotherapy was comprised of 5-fluorouracil, doxorubicin, and cyclophosphamide administered according to the following schedule: 5-fluorouracil at a dose of 500 mg/m2 intravenously (i.v.) on Days 1 and 8, doxorubicin at a dose of 50 mg/m2 as a continuous i.v. infusion over 72 hours through a central venous catheter on Days 1–3, and cyclophosphamide at a dose of 500 mg/m2 i.v. on Day 1. MV chemotherapy was comprised of the following: vinblastine at a dose of 1.7 mg/m2 as a continuous infusion daily on Days 1–5 and methotrexate at a dose of 75 mg/m2 i.v. on Day 1 followed by leucovorin calcium rescue given at a dose of 8 mg/m2 orally every 6 hours for 8 doses to begin 24 hours after methotrexate. The maximum total dose of doxorubicin was 300 mg/m2. The cycles of chemotherapy were repeated every 3–4 weeks depending on recovery of myelosuppression. Those patients randomized to the tamoxifen treatment arm received 20 mg daily for 5 years.

Dose Modification

The dosages of myelosuppressive drugs (doxorubicin and cyclophosphamide) were reduced by 20% in subsequent cycles if the lowest absolute neutrophil count was < 500/mL or the lowest platelet count was < 50,000/mL in the previous cycle, or if the patient had infectious complications regardless of the degree of neutropenia. For patients with > Grade 2 mucositis, the second dose of 5-fluorouracil (Day 8) was administered on the completion of the doxorubicin infusion (Day 3) in subsequent cycles. If that modification was inadequate in decreasing mucositis, then the duration of doxorubicin infusion was reduced to 48 hours during the next cycles.

Statistical Methods

The disease-free survival (DFS) and overall survival (OS) curves were estimated by the Kaplan–Meier method13 and comparisons were based on log-rank tests.14 The study was powered to detect an 8% difference in OS between the 2 treatment groups at a 5% significance level (power of test was 80%).

RESULTS

Between 1986–1994, 791 patients were enrolled in the current study. Two patients were ineligible (because of metastatic disease). A total of 789 eligible patients were followed for a median follow-up of 10 years since randomization. A total of 668 patients were randomized to receive either FAC or FAC + MV (Group 1). An additional 121 patients who were age ≥ 50 years with ER-positive tumors were randomized to receive tamoxifen or FAC + MV (Group 2). Patient characteristics for the two groups are shown in Tables 1 and 2, and respectively. There were no statistically significant differences with regard to patient characteristics. It is interesting to note that, in Group 1, approximately 50% of the patients had ER-negative disease and < 33% had lymph node-negative disease. It also is interesting to note that 30% of the patients were postmenopausal. Conversely, nearly all the patients in, in Group 2s had lymph node-positive disease. The results presented herein are on an intent-to-treat analysis.

Table 1. Characteristics of Patients in Group 1 Who Were Randomly Assigned FAC or FAC + MV Therapy
FactorTotalFAC (%)FAC + MV (%)P value
  1. FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine; ER: estrogen receptor; PR: progesterone receptor; Pre: premenopausal; post: postmenopausal.

Total668331337 
Race    
 White478243 (74)235 (70)0.28
 Hispanic105 43 (13) 62 (18)
 Black59 31 (9) 28 (8)
 Other26 14 (4) 12 (4)
ER    
 Negative350182 (55)168 (50) 
 Positive224 98 (30)126 (37)0.10
 Unknown94 51 (15) 43 (13) 
PR    
 Negative323161 (49)162 (49) 
 Positive237116 (35)121 (36)0.92
 Unknown103 53 (16) 50 (15) 
Stage    
 178 36 (11) 42 (12)0.42
 2510251 (76)259 (77)
 3A64 33 (10) 31 (9)
 3B16 11 (3)  5 (2)
Lymph node    
 0188 86 (26)102 (30) 
 1–3290151 (46)139 (41)0.40
 ≥ 4190 94 (28) 96 (29) 
Tumor size (cm)    
 < 3404204 (62)200 (59) 
 3–5186 88 (27) 98 (29)0.77
 ≥ 577 38 (11) 39 (12) 
Menopause    
 Pre468235 (72)233 (70)0.59
 Post194 93 (28)101 (30)
Table 2. Characteristics of Patients in Group 2 Who Were Randomly Assigned FAC + MV or Tamoxifen Therapy
FactorTotalTamoxifen (%)FAC + MV (%)P value
  1. FAC + MV: 5-fluorouracil, doxorubicin, and cyclophos-phamide plus methotrexate and vinblastine; PR: progesterone receptor; Pre: premenopausal; post: postmenopausal.

Total1215962 
Race    
 White9446 (78)48 (77)0.97
 Hispanic13 7 (12) 6 (10)
 Black13 6 (10) 7 (11)
 Other1 0 (0) 1 (2)
PR    
 Negative2813 (22)15 (24) 
 Positive8943 (74)46 (74)0.87
 Unknown3 2 (4) 1 (2) 
Stage    
 I5 5 (9) 00.12
 II10448 (81)56 (90)
 IIIA6 3 (5) 3 (5)
 IIIB6 3 (5) 3 (5)
Lymph node    
 01410 (17) 4 (6) 
 1–35925 (42)34 (55)0.14
 ≥ 44824 (41)24 (39) 
Tumor size    
 < 37635 (60)41 (66) 
 3–53519 (33)16 (26)0.74
 ≥ 59 4 (7) 5 (8) 
Menopause    
 Pre6 2 (4) 4 (9)0.68
 Post8443 (96)41 (91)

FAC versus FAC plus MV

In Group 1, 331 patients were randomized to receive FAC and 337 were randomized to receive FAC plus MV. The numbers of recurrences in each treatment arm were 120 and 110, respectively. There were 92 deaths in the FAC arm and 85 deaths in the FAC + MV arm. The estimated 5-year DFS estimates were 0.70 and 0.76, respectively (P = 0.26) (Fig. 1). The 5-year OS estimates were 0.84 and 0.83, respectively, for the FAC and the FAC + MV arms (Fig. 2).

Figure 1.

Disease-free survival by assigned treatment. FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine.

Figure 2.

Overall survival by assigned treatment. FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine.

Because the FAC and FAC + MV regimens were assigned randomly according to patient age and ER status, the comparison of DFS for the two treatment arms was considered also within subsets defined by patient age and ER status. A slight advantage for the FAC + MV regimen was noted in both age groups, suggesting there was no significant differential treatment effect by age (Fig. 3). However, when patients were grouped by ER status, it was notable that the DFS superiority of the FAC + MV regimen was observed only in the ER-positive group, although the difference did not reach statistical significance (Fig. 4). Because of the method of treatment assignment, all ER-positive patients in Group 1 were age < 50 years. However, it should be noted that 85 of the 337 patients randomized to receive FAC + MV (25%) received only FAC and 2 of the 337 patients received other therapies. Conversely, of those patients randomized to receive FAC, three received FAC + MV, one patient received no treatment, and one patient received alternative therapy. Analysis of DFS and OS according to actual treatment received did not demonstrate any survival advantage for the FAC + MV group (data not shown).

Figure 3.

Disease-free survival, grouped by patient age. FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine.

Figure 4.

Disease-free survival for patients grouped by estrogen receptor status. POS: positive; FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine; NEG: negative; UNK: unknown.

FAC + MV versus Tamoxifen

Patients age > 50 years with ER-positive disease were randomly assigned to receive FAC + MV or tamoxifen. As shown in Table 2, the characteristics of the patients did not appear to differ significantly between treatment groups. However, the small numbers of patients whose disease was Stage I or who had lymph node-negative disease were more likely to be assigned to receive tamoxifen.

Among 121 patients in Group 2, 59 were assigned to receive tamoxifen and 62 were assigned to receive FAC + MV. On the tamoxifen arm, 24 recurrences and 23 deaths occurred. At 5 years, 78% of patients in the tamoxifen arm were alive and free of disease versus 66% of patients in the FAC + MV group (P = 0.28) (Fig. 5). Overall survival at 5 years was 85% in the tamoxifen arm and 74% in the chemotherapy arm (P = 0.86) (Fig. 6).

Figure 5.

Disease-free survival by assigned treatment. TAM: tamoxifen; FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine.

Figure 6.

Overall survival by assigned treatment. TAM: tamoxifen; FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine.

Hematologic toxicities from these therapies have been reported in earlier publications.11, 15 However, all patients were monitored for any Grade 3 or 4 toxicity (NIH/NCI CTC version 1) and data are available for mucositis, neuropathy, sepsis, and catheter complications (Table 3). A total of 331 patients were randomized to receive FAC and 399 were randomized to receive FAC + MV (Groups 1 and 2). No major complications were noted in the patients treated with tamoxifen. The most common toxicity was mucositis; Grade 1-2 mucositis was reported in 11% and 15% of patients, respectively, in the FAC and FAC + MV groups. Grade 3–4 mucositis was reported in 7% of patients in the FAC arm and 8% of patients in the FAC + MM arm. Patients treated with FAC + MV also developed neuropathy (Grade 1–2: 3% and Grade 3–4: 2%) (Table 3). Sepsis was noted in 2% of the patients in each chemotherapy arm and clinical congestive heart failure was noted in < 1% of both groups. Finally, catheter complications (infection, thrombosis, pneumothorax, and hemothorax) were noted in 5% of patients in the FAC arm and 10% of patients in the FAC + MV arm.

Table 3. Toxicity Data
 MucositisNeuropathyCHFSepsisCatheter complicationsa
Grade 1–2Grade 3–4Grade 1–2Grade 3–4
  • CHF: congestive heart failure; FAC: 5-fluorouracil, doxorubicin, and cyclophosphamide; MV: methotrexate and vinblastine.

  • a

    Catheter complications were comprised of infection, thrombosis, and pneumothorax/hemothorax.

FAC (n = 331)37 (11%)23 (7%)002 (<1%)6 (2%)16 (5%)
FAC + MV (n = 399)61 (15%)34 (9%)13 (3%)9 (2%)2 (<1%)8 (2%)40 (10%)

DISCUSSION

This trial, launched in 1986, attempted to explore the merit of additional chemotherapy after six cycles of anthracycline-based chemotherapy for women with operable breast carcinoma. The rationale for the current study was based on the hypothesis that women who are more likely to develop a disease recurrence may benefit from more extended chemotherapy comprised of drugs with different mechanisms of action and no or minimal cross-resistance. Because, at the time, the standard of care for postmenopausal women with ER-positive breast carcinoma was hormonal therapy alone (tamoxifen), in that specific setting we compared “maximum” chemotherapy (FAC + MV) versus tamoxifen. We chose age as the cutoff value (age > or < 50 years) versus the more commonly used premenopausal/postmenopausal status.

At the M. D. Anderson Cancer Center, we have been using doxorubicin-based regimens in the adjuvant setting for > 25 years. In fact, six cycles of FAC has been our standard for adjuvant chemotherapy for patients with operable breast carcinoma for many years. To that we added four cycles of a combination of vinblastine and methotrexate, two drugs that have shown activity in the metastatic setting and have partial lack of cross-resistance with doxorubicin.8, 15

We did not find any significant difference with regard to the DFS or OS among the two groups receiving FAC and FAC + MV. Of all the studies that have addressed this question, to our knowledge the current series has the longest reported follow-up published to date (10 years). It is interesting to note that the subgroup of patients who were age > 50 years and had ER-positive disease demonstrated a trend toward improved DFS and OS when treated with tamoxifen alone for 5 years as opposed to chemotherapy with FAC + MV administered for 10 cycles (although the trend did not reach statistical significance).

To our knowledge, the current study is not the first to report the superiority of tamoxifen alone compared with polychemotherapy. The GROCTA trial group found tamoxifen to be superior to six courses of CMF followed by four courses of epirubicin in lymph node-positive patients, especially for postmenopausal women.16 Hubay et al. reported that 40 mg of daily tamoxifen for 1 year led to improved DFS compared with oral CMF for 12 months.17 In addition, there have been studies demonstrating no survival benefit from adding polychemotherapy to tamoxifen.18–20 However, there are a number of randomized trials that reported the superiority of adding chemotherapy to tamoxifen in patients with both lymph node-positive21, 22 and lymph node-negative23 disease. In fact, the overview analysis3 settled the issue by showing that by adding polychemotherapy to tamoxifen, patients derived an additional relative risk reduction of 11% for women age < 50 years to 25% for women age > 50 years, leading to the current standard of combined chemotherapy and hormonal therapy for all patients with hormone-positive disease.

An interesting observation in the current study was that DFS was prolonged with the FAC + MV regimen compared with the. FAC alone in the subgroup of patients with ER-positive disease in Group 1. Because these all were patients who were age < 50 years, the improved DFS with more extensive treatment could be the result of young age and not ER status. However, when we analyzed patients age < 50 years with negative ER status there was no difference in the DFS between the FAC and FAC + MV treatment arms.

As stated earlier, taxanes in sequence with doxorubicin-based regimens currently are being investigated in randomized, large-scale Phase III trials. Treatment with paclitaxel given for four cycles after four cycles of doxorubicin and cyclophosphamide (AC) has to date produced conflicting results. The Cancer and Leukemia Group B (CALGB) trial5 of sequential chemotherapy for lymph node-positive breast carcinoma demonstrated prolonged DFS (86% for AC alone vs. 90% for AC followed by paclitaxel; P = 0.0077) and OS (95% for AC alone vs. 97% for AC followed by paclitaxel; P = 0.039) with the addition of paclitaxel. However, further follow-up of that trial, as well as a similar trial by the National Surgical Adjuvant Breast and Bowel Project (NSABP) group, suggested that the addition of paclitaxel to four cycles of AC might be of benefit only to patients with ER-negative disease.1 A smaller trial from the M. D. Anderson Cancer Center randomized patients with operable breast carcinoma to receive eight cycles of FAC vs. four cycles of paclitaxel followed by four cycles of FAC.6 The question posed by this trial was whether the initial positive results reported with the addition of paclitaxel to an anthracycline-based regimen5 were the result of merely more chemotherapy as opposed to a specific therapeutic effect of paclitaxel. Preliminary results demonstrated a trend toward prolonged DFS at a median follow-up of 4years (83% for FAC ×8 vs. 86% for paclitaxel ×4 followed by FAC ×4), which did not reach statistical significance (P = 0.09).6

The role of docetaxel in this setting also has been investigated and preliminary results were presented recently.4 The addition of 4 cycles of docetaxel to 4 cycles of AC, either preoperatively or postoperatively, led to a significant increase in the rates of pathologic complete responses (pCR) in the primary tumor (13.7% for AC vs. 25.6% for AC + docetaxel), albeit at the expense of significant Grade3 or 4 toxicity (23% for AC vs. 40% for AC + docetaxel). Although to our knowledge no data concerning DFS or OS were presented (the follow-up period for the study is too short), it is postulated that increased pCR rates will translate into improved disease control.

The current study, which to our knowledge has the longest follow-up reported to date, failed to demonstrate any additional benefit from adding potentially noncross-resistant chemotherapy to six cycles of anthracycline-based chemotherapy. This could be the result of the choice of agents (i.e., vinblastine and methotrexate) that are inferior to taxanes. However, even with the adjunct of taxanes, it remains to be determined whether adding to anthracycline-based regimens confers long-term benefits and if so, in what subgroups of patients (CALGB, National Institutes of Health consensus).

Ancillary