- © 2001 by American Society of Clinical Oncology
Dose-Intensive Epirubicin-Based Chemotherapy Is Superior to an Intensive Intravenous Cyclophosphamide, Methotrexate, and Fluorouracil Regimen in Metastatic Breast Cancer: A Randomized Multinational Study
- Stephen P. Ackland,
- Antonio Anton,
- George Peter Breitbach,
- Elena Colajori,
- Jennifer M. Tursi,
- Carlos Delfino,
- Anna Efremidis,
- Adnan Ezzat,
- Alberto Fittipaldo,
- Kristo Kolaric†,
- Massimo Lopez,
- Donatella Viaro and
- the HEPI 013 Study Group
- From the Department of Medical Oncology, Newcastle Mater Misericordiae Hospital, Waratah, Australia.
- Address reprint requests to Stephen Ackland, MD, Department of Medical Oncology, Newcastle Mater Misericordiae Hospital, Locked Bag 7, Hunter Region Mail Center, 2310 Australia; email: mdspa@ alinga.newcastle.edu.au.
Abstract
PURPOSE: To determine the relative efficacy of a cyclophosphamide epirubicin and fluorouracil (CEF) regimen compared with an intravenous (IV) cyclophosphamide, methotrexate, and fluorouracil (CMF) combination in metastatic breast cancer.
PATIENTS AND METHODS: Patients were randomized to receive either CEF (cyclophosphamide 400 mg/m2 IV, epirubicin 50 mg/m2 IV, and fluorouracil 500 mg/m2 IV on days 1 and 8), or CMF (cyclophosphamide 500 mg/m2 IV, methotrexate 40 mg/m2 IV, and fluorouracil 600 mg/m2 IV on days 1 and 8). Treatment was given in 3- to 4-week cycles for a total of six to nine cycles.
RESULTS: A total of 460 patients (223 CEF and 237 CMF) were randomized. Overall response rate was superior for CEF than CMF in all randomized patients (57% v 46%, respectively; P = .01) and in the assessable subset (66% v 52%, respectively; P = .005). With a median follow-up of more than 20 months, time to progression (TTP) was significantly longer with CEF than CMF (median 8.9 v 6.3 months, respectively; P = .0064), as was time to treatment failure (TTF) (median 6.2 v 5.0 months, respectively; P = .01). Significant survival differences were not observed between CEF and CMF (median 20.1 v 18.2 months, respectively; P = .23). Granulocytopenia and infections were similar in both arms. Grade 3/4 nausea/vomiting and alopecia were more frequent with CEF, whereas diarrhea was more frequent with CMF. Cardiac toxicity, primarily asymptomatic, required withdrawal from study of 15 patients on CEF (7%) and one patient on CMF.
CONCLUSION: This CEF regimen safely provides significantly better tumor control than CMF, manifest as a higher response rate, and longer TTP and TTF, but not survival, when used as first-line chemotherapy for metastatic breast cancer.
BREAST CARCINOMA is the most common cancer in women (30% of all cancers) and is second only to lung cancer as a cause of cancer deaths.1 The median survival time from diagnosis of metastases is approximately 3 years.2 Systemic treatment with hormonal therapy or chemotherapy is of significant palliative benefit in patients with metastatic disease.
Various cyclophosphamide, methotrexate, and fluorouracil (CMF) combinations have traditionally been used in the treatment of metastatic breast cancer. The classical schedule of CMF comprises oral cyclophosphamide (CTX) 100 mg/m2 on days 1 to 14, with intravenous (IV) methotrexate 40 mg/m2 and fluorouracil (5-FU) 600 mg/m2 given on days 1 and 8 cycled every 28 days. Although this is generally thought to be well tolerated, the gastric disturbance associated with the use of oral cyclophosphamide is frequent and accounts for poor compliance in 30% to 40% of patients.3,4 CMF combinations using IV CTX have been used to overcome the problems of nausea, vomiting, and attendant noncompliance, but there has been concern as to whether the IV CMF combinations are equally efficacious as the oral CMF regimen. In a large European Organization for Research and Treatment of Cancer study involving 254 postmenopausal patients with metastatic breast cancer, the classical oral CMF regimen had a significantly higher response rate and longer time to tumor progression and survival than IV CMF.5 However, improvements in efficacy were offset by significantly worse toxicity (hematologic, mucositis, and alopecia). The higher dose-intensity of CTX with classical oral CMF (1.0 v 0.62) may have been the major contributor to the differences in antitumor activity.
The anthracyclines are among the most active chemotherapeutic agents in the treatment of metastatic breast cancer and now form a part of many adjuvant and palliative chemotherapy regimens.6-8 Oral and IV CMF regimens have been compared to doxorubicin-containing regimens in a number of studies in metastatic breast cancer9,10 where the role of anthracyclines in first-line treatment has been confirmed with significantly better response rates but varying effect on time-related parameters and generally no survival benefit in individual studies. An overview of five studies totaling 1,088 patients showed a survival benefit for doxorubicin-containing regimens compared with CMF regimens (hazard ratio 0.78, 95% confidence interval [CI], 0.66 to 0.90).11
Epirubicin, an analog of doxorubicin, has been shown to have less acute hematologic and nonhematologic toxicity than doxorubicin when used as a single agent and in combination with other agents.12-17 The dose and dose-intensity of the administered chemotherapy are important variables that can be adjusted in attempting to improve response in patients with breast cancer. In particular, dose-intensity analyses of anthracyclines in breast cancer have shown a positive and significant correlation between higher doses and response.18,19 Epirubicin is an ideal candidate for dose intensification in patients with breast cancer because of its anticancer activity and lower overall toxicity compared with doxorubicin.15,16 Epirubicin shows a dose-response relationship when used as a single agent20-23 or in combination chemotherapy.24-28 These findings suggest that dose intensification of epirubicin is feasible and may lead to better outcomes in patients with metastatic breast cancer.
Against this background, we have undertaken a randomized controlled trial comparing a dose-intensive epirubicin-based combination with an IV CMF regimen with a similar dose-intensity of CTX as that of classical oral CMF.29 The end points for the study were time to progression, response rate, time to treatment failure, survival, and toxicity.
PATIENTS AND METHODS
Eligibility criteria
Eligible patients were females between 18 and 70 years of age with histologically proven breast cancer and metastases which were measurable and/or assessable by standard World Health Organization (WHO) criteria.30 Patients had no prior chemotherapy for metastatic breast cancer and no prior adjuvant anthracycline or anthracenedione. No more than two regimens of hormonal therapy for metastatic disease were allowed. Patients were required to have a WHO performance status of 0, 1, or 2; adequate marrow function (WBC count ≥ 4 × 109/L, platelet count ≥ 100 × 109/L unless there was documented evidence of bone marrow involvement), adequate liver function (total bilirubin < 1.5 mg/dL or 25 μmol/L and AST < 3 × normal unless AST abnormalities were related to liver metastases), and adequate renal function (creatinine < 1.5 mg/dL or < 130 μmol/L). A resting left ventricular ejection fraction (LVEF) within normal limits of the institution was required. Patients were not eligible if they had failed adjuvant treatment less than 12 months previously, had CNS involvement, inflammatory carcinoma of the breast, or other significant malignancies (history of other malignancy other than localized basal cell or squamous cell skin carcinoma or noninvasive cervix carcinoma < 5 years before entering the study), an active infectious process, or were pregnant. Patients with significant cardiovascular history or ECG abnormalities were not eligible. All patients gave written or oral witnessed informed consent according to local Institutional Ethics Committee guidelines.
Treatment Plan
Patients were stratified within each institution according to the presence or absence of visceral metastases, by the number of metastatic sites involved (one to two or > two organs involved), and whether they had received previous adjuvant chemotherapy or not. Thus a total of eight strata were identified and a separate randomization code was computer generated and centrally assigned for each stratum in each center.
Patients were randomized to receive either CEF (cyclophosphamide 400 mg/m2 IV, epirubicin 50 mg/m2 IV, and 5-FU 500 mg/m2 IV each on days 1 and 8) or CMF (cyclophosphamide 500 mg/m2 IV, methotrexate 40 mg/m2 IV, and 5-FU 600 mg/m2 IV each on days 1 and 8), with the cycles repeated every 3 or 4 weeks depending on recovery from toxicity. To calculate actual dose, the lesser of ideal and actual body weight was used.
Dose modifications were based on weekly complete blood count results and a thrice weekly assessment of other toxicities. On day 8, doses of CTX and 5-FU were reduced to 75% if the absolute neutrophil count (ANC) was ≤ 1.0 × 109/L, and omitted if the platelet nadir was also less than 100 × 109/L. On subsequent cycles, dose reduction to 75% of the previous cycle occurred if the ANC nadir was less than 0.2 and/or platelet nadir ≤ 50 × 109/L, or if febrile neutropenia or other grade 3/4 toxicity had occurred. On day 21, treatment was delayed if the ANC was less than 1.5 × 109/L or platelets less than 100 × 109/L. A maximum delay of 2 weeks was permitted for the recovery of toxicity parameters after which the patient went off study treatment as a result of toxicity.
Treatment continued until six cycles had been given (nine cycles if complete or partial response was achieved), unless disease progression (PD), toxicity, or patient refusal occurred earlier. Patients who had stable disease as best response could only receive another treatment after PD had occurred.
Other treatment
No concomitant chemotherapy or hormonal therapy was allowed from the initiation of the study treatment until progressive disease was recorded. Corticosteroids were allowed as part of antiemetic therapy. Febrile neutropenia was treated according to institutional policy. When progression of disease occurred patients were treated according to physician choice.
Schedule of Evaluations
At registration all patients had a history, complete physical examination, WHO performance status, complete blood cell count (CBC), serum chemistry studies (electrolytes, urea, creatinine, calcium, bilirubin, AST, and alkaline phosphatase), ECG, and LVEF (by radionuclide gated heart scan or echocardiography). Tumor evaluation was undertaken with clinical examination, chest x-ray, liver computed tomography or ultrasound, bone scan, and bone x-rays if the scan was positive.
Physical examination, performance status, chemistry, and ECG were repeated every 3 weeks. CBC was performed weekly, clinical toxicity was measured at baseline and at weeks 1 and 3 of each cycle. All evaluations and LVEF (radionuclide or echocardiography) were repeated at the end of therapy. In the CEF arm LVEF was also measured when epirubicin cumulative doses of 400 to 500mg/m2 and 700 to 800 mg/m2 were reached and before each cycle thereafter. The criteria for stopping study treatment because of cardiotoxicity were a ≥ 20% absolute decrease from baseline, or a more than 10% decrease below the lower limit of normal as defined by the institution, or clinical congestive heart failure. Tumor assessment was performed every 6 weeks. After treatment completion, patients were followed for PD and survival every 3 months.
Efficacy Analysis
Patient eligibility, assessability, and best response achieved were confirmed by a review of the case report forms by an independent medical oncologist. A patient was considered assessable for response if she had received at least two cycles of treatment, unless rapid PD occurred earlier, in which case treatment was considered to have failed. WHO criteria for response were used.30 All assessments of all lesions were used to determine overall response. Stabilization of disease for at least 4 months, rather than 1 month, was required to qualify for the no change category.
Efficacy analyses was performed on all patients on an intent-to-treat (ITT) basis. Parameters assessed were time to progression (TTP), time to treatment failure (TTF), overall survival, and response rate. Response rate was analyzed on an ITT basis in all randomized patients who on review had histologically proven breast cancer. In addition, response rate, TTP, and duration of response were analyzed in all randomized patients who were eligible and assessable for efficacy assessment and were treated as randomized.
Safety Analysis
All patients who had received at least one cycle of chemotherapy were included in the safety analysis according to treatment actually received. All toxicities were evaluated at each cycle by the WHO criteria. In the case of hemoglobin at least one evaluation during the cycle was required. For other hematologic parameters, a cycle was defined as assessable if at least one CBC was available between days 8 and 15 (inclusive). Cardiac toxicity was evaluated according to the cumulative dose of epirubicin administered. LVEF analysis was performed in all assessable patients (defined as those patients with a baseline and at least one subsequent assessment using the same method).
Statistical considerations
The primary end point of this study, TTP, was defined as the time in days between the date of randomization and the first date of documented PD or death resulting from any cause, whichever occurred first. If the patient received fewer than six cycles of therapy and the reason for off treatment was refusal, toxicity/adverse event or loss to follow-up, the last date of clinical or laboratory evaluation was used to censor the patient. Secondary variables of interest were response rate, calculated for each treatment as the ratio between patients classified as complete response + partial response and all randomized patients, and TTF, defined as the interval in days from the date of randomization to the first date of treatment failure (disease progression, death, treatment discontinuation resulting from toxicity, patient refusal, or loss to follow-up). Survival was defined as the time in days from the date of randomization to the date of death.
The primary objective of the study was to detect an advantage in TTP for the CEF arm as compared with the CMF arm (11 months v 8 months). Overall, 155 events were required to provide 80% power to the analysis, at a significance level of 5% (two-tailed test). Based on the assumptions of an accrual time of 15 months and performance of the analysis 12 months after the accrual completion, 210 patients per arm were estimated.31
Response rates were compared using a χ2 test. Time-dependent variables (TTP, TTF, and overall survival) were estimated using the Kaplan-Meier method32 and compared using the log-rank test. Additionally, for TTP a Cox proportional hazard regression analysis was performed to investigate the effect of the stratification factors on outcome.33
For safety analysis, toxicity scores were analyzed on both per patient and per cycle bases. Graded scores were compared by a χ2 test between the two treatment arms. Changes in LVEF were calculated as the difference between baseline and each subsequent measurement.
Dose-Intensity
Dose-intensity (DI), relative DI, and average relative DI were calculated according to Hryniuk and Bush.18 For each drug administered, DI expressed as mg/m2/wk was calculated by summing each cycle dose in mg/m2 divided by the number of weeks from the first day of cycle 1 to the date of the last cycle plus a fixed time of 3 weeks. The relative DI was calculated as the ratio between the DI for each drug as received and the planned DI. The average relative DI was calculated as the mean relative DI for each drug.
RESULTS
Definition of study population
Between September 1990 and November 1992, 460 patients (223 CEF, 237 CMF) were entered from 48 centers in 20 countries ( Fig 1). The number of patients registered from single centers ranged from one to 28. One additional patient was registered to the CEF arm but had begun treatment before randomization, and therefore is not included among the 223 CEF patients. Of the 223 patients randomized to the CEF arm, 218 were treated and five patients were never treated (two patient refusals, one protocol violation, one lost to follow-up, one treated by another physician after randomization). Of the 237 patients randomized to the CMF arm, one patient was not treated and two patients were given CEF in error. After review of the eligibility criteria, 217 of the 223 patients in the CEF group were found to be eligible; six patients were not eligible (three abnormal baseline parameters, two prior therapy, and one brain metastases). Of these 217 patients, 189 patients (85%) were considered assessable for response. In the CMF group, 224 (95%) patients were found to be eligible; 13 patients were not eligible (one wrong diagnosis, five unconfirmed diagnosis, five prior therapy, one brain metastases, and one concomitant treatment). Of the 224 patients, 200 patients (84%) were considered to be assessable.
Fig 1. Flow chart of patients registered on study.
The distributions of patients’ characteristics for the ITT population in the two trial arms are listed in Table 1. Baseline characteristics of age, menopausal status, performance status, estrogen receptor status, metastatic patterns, and prior treatments were well balanced between the two arms. The majority of patients had symptoms related to breast cancer as evidenced by a baseline performance status of 1 or 2 in 66% of CEF-treated patients and 64% of CMF-treated patients. Approximately one quarter of the patients in each treatment group had metastatic disease at diagnosis. More than half of the CEF and CMF patients (57% and 60%, respectively) had visceral disease.
Patient Characteristics
Prior treatment was well balanced between the two arms. Thirty-four percent of patients allocated to CEF and 33% allocated to CMF had received prior adjuvant hormonal therapy. Similarly, 20% and 22%, respectively, had received adjuvant chemotherapy. In both arms, 31% of patients had disease confined to one organ, with 33% (CEF) and 35% (CMF) having disease in three or more organs.
Treatment administration
Information regarding treatment administration and the median relative DI is listed in Table 2. In both treatment groups, the median number of cycles administered was six. Reduction in doses and the lengthening of the planned 3-week cycling was sometimes necessary to accommodate individual patient tolerance. Primarily because of this modification in schedule, the median relative DIs of each of the drugs was 74% to 77% of the planned DIs. As planned by protocol the median actual dose intensities of CTX and 5-FU in the CEF arm of the study were lower than those in the CMF arm. The median cycle duration was 28 days for both treatment groups, with 17% of the CEF and 27% of the CMF treatment cycles of 3-weeks duration.
Treatment Administration
Efficacy Results
There was a significant difference in overall objective response rate in favor of CEF compared with CMF in all randomized patients with a diagnosis of breast cancer (57% v 45%, respectively, P = .007). A significantly higher response rate was also observed with CEF than CMF in the assessable patient population (66% v 52%, respectively, P = .005) ( Table 3) with the odds ratio for response to CEF versus CMF of 1.8 (95% CI, 1.2 to 2.7). More patients treated with CEF had complete responses in both analyses, although the difference was not significant (14% CEF v 11% CMF by ITT and 16% CEF v 13% CMF in assessable patients, respectively). A total of 259 patients (134 CEF, 125 CMF) received at least six cycles of therapy. Of those, 108 CEF and 92 CMF achieved either a complete response or partial response as best response as determined by the independent reviewer. In both therapies, 91% of responders achieved their best response within the sixth cycle of therapy. Among the responding patients the average length of observation was similar in the two therapy arms (21 and 19 months for CEF and CMF, respectively) with only one patient lost to follow-up in the CMF group.
Antitumor Response
The response rate was significantly higher for CEF in the subset of patients with nonvisceral disease (78% CEF v 58% CMF, P = .005) (Table 3) and those with more than two sites involved (65% CEF v 41% CMF, P = .006). There was a trend toward a longer median duration of response in the CEF group than the CMF group (8.8 months v 7.2 months, respectively; P = .07).
TTP was significantly longer in patients randomized to CEF than with CMF (median, 8.7 months v 6.3 months; P = .01) ( Fig 2). The hazard ratio for CEF versus CMF was 0.73 (95% CI, 0.59 to 0.92), with a relative reduction rate for progression (CEF v CMF) of 0.27 (95% CI, 0.08 to 0.41).
Fig 2. Kaplan-Meier curve for TTP (all randomized patients).
A Cox regression analysis was performed to evaluate the effect of putative baseline prognostic factors on TTP. The final multivariate model included assigned treatment, dominant metastases (visceral versus nonvisceral), and number of sites (one to two v > two sites) confirming the significant effect of treatment on TTP (P = .004; hazard ratio [CEF/CMF] = 0.72). More than two sites of disease (hazard ratio = 1.7) and presence of visceral metastases (hazard ratio = 1.4) were confirmed as unfavorable prognostic factors for TTP. Prior adjuvant therapy did not significantly predict for TTP.
The time to treatment failure significantly favored the CEF arm (median 6.2 months, 95% CI, 5.7 to 6.7) compared with CMF (median, 5.0 months; 95% CI, 4.3 to 5.4) (P = .01) ( Fig 3). The hazard ratio for TTF (ITT analysis) on randomized patients with histologically proven breast cancer for CEF versus CMF was 0.77.
Fig 3. Kaplan-Meier curve for TTF (all randomized patients).
The median survival was 20.1 months (95% CI, 18 to 23) in patients treated with CEF compared with 18.2 months (95% CI, 17 to 21) in CMF-treated patients ( Fig 4). The hazard ratio of CEF versus CMF for survival was 0.87 (95% CI, 0.7 to 1.1), and the difference in survival was not significant (P = .23). It should be noted that anthracycline treatment as second-line therapy was administered to 44% of patients failing CMF treatment as compared with 18% of CEF patients.
Fig 4. Kaplan-Meier curve for overall survival (all randomized patients).
Toxicity
Hematologic toxicity, consisting mainly of granulocytopenia and leucopenia, was the most frequent acute toxicity in both treatment groups ( Table 4). The median nadir for all hematology parameters was generally slightly lower in the CEF group than the CMF group, and grade 4 neutropenia was somewhat more frequent with CEF than with CMF (43% v 33%, respectively). However, we observed a similar incidence of febrile neutropenia (11% of CEF patients and 8% of CMF patients) and grade 4 infections (0.5% of CEF and 1% of CMF patients).
Hematologic Toxicities
The two regimens were generally similar in terms of the incidence of nonhematologic toxicity ( Table 5). Although grade 3-4 nausea and vomiting was somewhat more frequent in the CEF group than in the CMF group (21% v 14% of the patients and 6% v 4% of cycles, respectively), only 1% of CEF patients experienced grade 4 vomiting. Diarrhea of any grade was more common in CMF-treated patients (22%) than in those receiving CEF (16%). As was expected for an anthracycline-containing regimen, CEF caused significantly more alopecia, which was of grade 3-4 in 66% of CEF-treated patients versus 14% of CMF-treated patients (P < .05). No clinically significant adverse effect of either treatment on liver or renal function was evident.
Grade 3-4 Nonhematologic Toxicities (as-treated population)
Of patients treated with CEF, 156 (71%) were assessable for changes in LVEF whereas only 72 (31%) of CMF patients were assessable. Nineteen patients given CEF had a significant fall in LVEF as defined earlier ( Table 6). The median cumulative dose of epirubicin administered in this study was 582 mg/m2 (range, 49 to 950 mg/m2). There was a trend toward an increasing risk of cardiotoxicity with increasing cumulative doses of epirubicin. Of the 15 patients who discontinued treatment as a result of adverse cardiac events, nine were because of declines in LVEF; six other patients were taken off study treatment because of ECG changes (three patients) and congestive heart failure (three patients). The latter three patients had total cumulative doses of epirubicin of 583 mg/m2, 600 mg/m2, and 688 mg/m2, respectively. One of these patients had underlying baseline cardiac arrhythmias, requiring medication. In addition, two other patients developed congestive heart failure after discontinuation of chemotherapy at cumulative epirubicin doses of 700 and 900 mg/m2. Six patients in the CEF treatment arm also presented with signs and symptoms consistent with the known early cardiac toxicity of anthracyclines such as tachycardia and other ECG abnormalities. These effects are usually not predictive of subsequent development of delayed cardiomyopathy. In all patients the events were not of clinical importance; most did not require any therapeutic intervention and did not necessitate the suspension of the anthracycline treatment. Two patients treated with CMF had significant falls in LVEF after six and seven cycles.
Cumulative Cardiotoxicity
Twenty-seven (12%) CEF patients discontinued treatment as a result of adverse events, which were cardiac in 15 (7%) and other toxicities in 12 (5%). Of the CMF-treated patients, 10 (4%) discontinued treatment because of toxicity; one (0.4%) had cardiac toxicity. Overall, there were seven toxic deaths in the CEF group (three possibly treatment-related) and nine in the CMF group (two possibly treatment-related), but none of these were cardiac events.
DISCUSSION
This study demonstrates that a dose-intensive CEF regimen is significantly better in several outcome measures than an IV CMF regimen of similar DI as first-line chemotherapy of metastatic breast cancer. The primary objective of this study was to detect a difference in TTP, and this aim was met; TTP was significantly longer for CEF than CMF (median, 8.7 months v 6.3 months, respectively; P = .01). In addition, the response rate was significantly higher with CEF (57% v 45%, respectively; P = .007), and the TTF, an overall measure of efficacy plus toxicity, was significantly in favor of CEF (median, 6.2 months v 5.0 months, respectively; P = .01).
Chemotherapy in metastatic breast cancer is effective for disease palliation and improvement in quality of life but few studies have demonstrated significant differences in survival between regimens.11 This study is not dissimilar in that we observed a trend but no significant improvement in survival for CEF versus CMF in this study (median, CEF 20.1 months v CMF 18.2 months). Multiple factors other than initial treatment allocation, including second-line treatments, may contribute to net outcome in terms of overall survival for patients with advanced breast cancer. No control was exercised over the treatment patients received after tumor progression in our study, with 44% of patients originally treated with CMF receiving second-line anthracycline therapy. However, other studies suggest that salvage chemotherapy after anthracycline failure does not substantially alter outcome. Second-line treatments are generally associated with low response rates and short duration of response.34 Newer second-line chemotherapy strategies, such as use of taxanes, vinorelbine, or dose-intensive chemotherapy, may in the future obscure survival gains from first-line chemotherapy.
This finding is typical of most randomized studies comparing different chemotherapy regimens for metastatic breast cancer, where an improvement in response rate for one treatment does not convert into improved survival. However, when dose-intensive epirubicin-containing regimens have been used in the adjuvant setting, the benefit does translate into significant survival advantages. A randomized study comparing a dose-intensive CEF regimen of cyclophosphamide 75 mg/m2 orally days 1 to 14, epirubicin 60 mg/m2 days 1 and 8, and 5-FU 500 mg/m2 days 1 and 8 every 4 weeks for 6 months with standard oral CMF as adjuvant therapy in 710 premenopausal patients with node-positive breast cancer demonstrated a better 5-year relapse-free survival (63% v 53%, respectively; P = .009) and 5-year survival (77% v 70%, respectively; P = .03) in patients receiving CEF.35 Also, in a study by Mouridsen et al,36 CEF proved to be superior to an IV CMF regimen in premenopausal women in terms of both relapse-free and overall survival based on a median 6-year follow-up.
The impact of prior adjuvant chemotherapy on study outcome is uncertain. Previous investigators have variously shown that adjuvant chemotherapy affects response, TTP, or overall survival after chemotherapy for advanced disease34,35,37,38, but others have shown no effect.39,40 Venturini et al41 studied 326 patients given a CEF regimen after prior adjuvant chemotherapy with CMF or anthracycline-based regimens and concluded that prior chemotherapy adversely affects outcome from CEF. In our study, only 21% of patients had prior adjuvant chemotherapy and in a multivariate analysis it was not a significant factor in outcome.
The optimal dose and schedule of CMF has been emphasized as necessary for achieving the best possible outcome in both the advanced and adjuvant settings.5,42 This was the rationale for the use of a day 1 and 8 schedule of CMF in the comparison with CEF. The planned DI of CTX in our CMF regimen was 333.3 mg/m2, which contrasts with the intended DI of only 200 mg/m2/wk of CTX in the IV CMF arm of the European Organization for Research and Treatment of Cancer study. Of note, the actual median delivered DIs of both CTX and 5-FU in the CEF arm of our study was lower than those in the CMF arm. The better outcomes that were observed in the CEF group despite the reduction in CTX and 5-FU delivery emphasizes the importance of epirubicin in the efficacy of the CEF regimen.
Dose and DI are considered important for optimal efficacy of chemotherapy, including epirubicin. Several studies have compared various doses of epirubicin as a single agent and suggest that higher doses are more effective.20-23,43 A dose-response relationship for epirubicin in combination with 5-FU and CTX has been shown in three studies, two in advanced/metastatic disease and another in early breast cancer.25,28,44 Focan et al25 compared two CEF regimens, one containing epirubicin 50 mg/m2 on both days 1 and 8 and the other containing 50 mg/m2 on day 1 only. The more dose-intensive regimen was superior in terms of response rate, TTF, and duration of response. A longer median survival, although not statistically significant, was observed with the higher dose CEF regimen compared with the lower dose regimen (27.1 months v 23.6 months, respectively). Brufman et al28 compared 5-FU 500 mg/m2 and cyclophosphamide 500 mg/m2 given with either epirubicin 50 mg/m2 (FEC50) or 100 mg/m2 (FEC100) IV every 3 weeks in patients with metastatic breast cancer. FEC100 was superior in terms of response rate, complete response rate, and response in visceral disease but TTP and overall survival were similar. In the adjuvant setting the French Adjuvant Study Group compared CEF100 and CEF50 regimens as adjuvant therapy in pre- and postmenopausal patients.44 Sixty-month follow-up data favor the higher DI treatment, demonstrating significantly longer relapse-free and overall survival. These data confirm the advantage of DI of chemotherapy including epirubicin in advanced breast cancer, at least in terms of higher response rate and better duration of response.
In our study, the two treatments had a similar incidence of acute toxicity (excluding alopecia) and the toxicity profile was consistent with that expected from these classes of drugs. The doses of CMF used produced similar myelosuppression to that of the CEF. Grade 4 neutropenia was observed more frequently with the CEF regimen but this did not lead to a significantly higher incidence of febrile neutropenia. Differences in cardiac events between the two treatments were as expected and account for a higher overall proportion of CEF patients stopping treatment because of toxicity. Fifteen patients in the CEF arm were withdrawn from treatment because of a decrease in left ventricular function or ECG changes compared with one patient in the CMF arm. However, only three of the 15 patients had heart failure and this occurred with cumulative epirubicin doses of 583, 600, and 688 mg/m2 respectively. Two other patients had posttreatment cardiac failure after doses of 700 mg/m2 and 900 mg/m2 respectively. The incidence of cumulative cardiotoxicity seen in this study (2%) is similar to other studies that utilized epirubicin in combination or as a single agent.15,16,28,45 These results are also consistent with the known relationship between the lifetime cumulative dose of anthracyclines and cardiotoxicity.46-48 In this context, epirubicin chemotherapy administered either alone or in combination offers a wider safety margin than doxorubicin.
In conclusion, this study shows that a dose-intensive CEF regimen administered as first-line chemotherapy of metastatic breast cancer has a significantly greater capacity to reduce tumor size and effect control than an intensive IV CMF combination, providing an improved response rate and longer TTP and TTF. In view of this increased tumor control, this dose and schedule of CEF provides an effective and safe regimen and constitutes a valid therapeutic option for patients with advanced breast cancer.
APPENDIX
The following HEPI-013 investigators participated in the study: Dr E. Abdi, Royal Adelaide Hospital, Department of Medical Oncology; Dr D. Kotasek, The Queen Elisabeth Hospital, Adelaide; Dr S. Ackland, Newcastle Mater Misericordiae Hospital, Department of Medical Oncology, Waratah; Dr G. Beadle, Wesley Clinic for Haematology & Oncology, Auchenflower; Dr M. Friedlander, Department of Oncology, Prince of Wales Hospital, Randwick, Australia; Dr A. Anton, Hospital Miguel Servet Servicio Oncologìa Médica, Zaragoza; Dr E. Aranda, Hospital Universitario “Reina Sofia,” Cordoba; Dr E. Murillo, Centro Regional e Oncologìa “Duque del Infantado;” Dr M. Nogueira, Ciudad Sanitaria “Virgen del Rocio,” Sevilla, Spain; Dr M. Balli, Ospedale Civile di Legnago, Reparto di Radioterapia e Oncologia, Verona; Dr M. De Lena, Ospedale Oncologico, Bari; Dr M. Lopez II, Divisione Oncologica Medica, Istituto Nazionale Tumori, Regina Elena, Roma; Dr G. Luporini, Ospedale San Carlo Borromeo, Divisione Oncologica, Milano; Dr E. Micheletti, Ospedali Civili Istituto del Radio, Brescia; Dr B. Morrica, Ospedale di Cremona, Divisione di Radioterapia Oncologica, Cremona; Dr M. Tordiglione, Ospedale di Circolo, Varese, Italy; Dr G.P. Breitbach, Stadtisches Krankenhaus, Neunkirchen; Dr W. Dornhoff, Krankenanstalt Muttherhaus der Borromaurinnen, Trier; Dr. C. Klinkenstein, Klinik fur Innere Medizin, Frankfurt/Oder; Dr C. Villena, Universitatskliniken im Landeskrankenhaus, HamburgFrauenklinik, Hamburg; Dr H. Wilken, Universitats Frauenklinik Rostock, Rostock, Germany; Dr J. Cervek, Institute of Oncology, Department of Chemotherapy, Ljubljana, Slovenia; Dr R. Chacòn, Istituto Fleming; Dr C. Delfino, Hospital Privado de la Comunidad; Dr R. Estévez, Instituto “Dr. Estévez;” Dr E. Mickiewicz, Hospital “Angel Roffo;” Dr H. Muro, Sanatorio Antàrtida, Buenos Aires, Argentina; Dr D. Damianov, National Oncological Centre, Sofia, Bulgaria; Dr D. Donat, Institute of Oncology, Sremska Kamenica, Novi Sad; Dr L. Vuletic, Institute of Oncology & Radiology, Department of Chemotherapy, Belgrade, Yugoslavia; Dr A. Efremidis, Hellenic Cancer Institute, St Saves Hospital, Athens, Greece; Dr A. Ezzat, King Faisal Specialist, Hospital & Research
APPENDIX (Cont’d)
Center, Riyadh, Saudi Arabia; Dr M.N.N. Gershanovich, Petrov Research Institute of Oncology, Therapy and Clinical Chemotherapy Departments, Russia; Dr N. Ghilezan, Institutul Oncologic, Cluj, Cluj, Rumania; Dr R. Hegg, Hospital das Clìnicas, Sao Paulo, Brasil; Dr J. Jassem, Medical Academy Oncological, Clinical Department of Radiotherapy, Gdansk; Dr M. Pawlicki, Oncological Institute, Department of Chemotherapy, Cracow; Dr C. Ramlau, Academy of Medicine, Poznan, Poznan; Dr P. Siedlecki, “Marie Sklodowka Curie” Oncological Institute, Department of Chemotherapy, Warsaw, Poland; Dr I. Kocak, “Masaryk” Memorial Cancer Centre Brno; Dr S. Korec, National Cancer Institute, Bratislava, Czechoslovak; Dr K. Kolaric, (now deceased) Institute for Tumors and Allied Diseases, Department of Chemotherapy, Zagreb, Croatia; Dr I.M. Muse, Hospital de Clinìcas, Montevideo, Uruguay; Dr T. Nagykàlnai, “Uzsoki” Hospital, Department of Oncoradiology, Budapest; Dr L. Perenyi, “Szent-Gyorgyi Albert” University Medical School, Radiological Clinic, Department of Radiology, Szeged; Dr T. Pinter, “Petz Aladar” Gyor County Hospital, Department of Oncoradiology, Gyor, Hungary; Dr C.S. Vallejos, Instituto Nacional de Enfermedades Neoplasicas del Perù, Lima, Perù; and Dr C. Vogel, South Florida Comprehensive Cancer Center, Miami, FL.
Acknowledgments
Supported in part by Farmitalia Carlo Erba SpA, Milano, Italy.
ACKNOWLEDGMENTS
We thank Prof Esteban Cvitkovic (Hopital Paul Brousse-SMST, Paris, France) for his expert advice in reviewing all CRFs and Dr Alessandro Riva for his contribution to the study. We are also indebted to all HEPI-013 participants for their constant collaboration on this project (all participating centers are listed in the Appendix).
- Received November 18, 1999.
- Accepted October 18, 2000.
