Journal of Cancer Therapy, 2011, 2, 342-353
doi:10.4236/jct.2011.23047 Published Online August 2011 (http://www.SciRP.org/journal/jct)
Randomized Trial Comparing Cyclophosphamide,
Methotrexate, and 5-Fluorouracil (CMF) Regimen
with Rotational CMFEV Regimen (E = Epirubicin,
V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma*
Giorgio Cocconi1, Corrado Boni2, Maurizio Tonato3, Rodolfo Passalacqua1, Mariantonietta Colozza3,
Anna M. Mosconi3, Giancarlo Bisagni2, Ermanno Rondini2, Lina Rodinò4, Amalia Carpi4,
Francesco Di Costanzo5, Mauro Brugia5, Giuseppe Attardo6, Luigi Acito7, Riccardo Rossetti8,
Maria Bella1, Roberta Camisa1, Francesco Cardinale9, Beatrice Dozin9
1Medical Oncology Division, Azienda Ospedaliera Universitaria, Parma, Italy; 2Medical Oncology Service, Azienda Ospedaliera,
Reggio Emilia, Italy; 3Medical Oncology Division, Azienda Ospedaliera Universitaria, Perugia, Italy; 4Medical Oncology Service,
Azienda Sanitaria Locale, Piacenza, Italy; 5Medical Oncology Service, Azienda Sanitaria Locale, Terni, Italy; 6Medical Oncology
Unit, Azienda Ospedaliera, Vigevano, Italy; 7Medical Oncology Division, Azienda Sanitaria Locale, Fermo, Italy; 8Miedical
Oncology Unit, Marciano, Italy; 9Epidemiologia Clinica, Istituto Nazionale per la Ricerca sul Cancro, Genova, Italy.
Received April 29th, 2011; revised June 2nd, 2011; accepted June 10th, 2011.
ABSTRACT
Objectives: The CMFEV (cyclophosphamide, methotrexate, 5-fluorouracil, epirubicin, vincristine) regimen is an inno-
vative schedule, designed by our Group, aimed at administering five partially or totally no cross-resistant cytotoxic
agents in breast carcinoma. It was randomly compared to CMF (cyclophosphamide, methotrexate, 5-fluorouracil) as
primary treatmen t in operab le disea se and demons trated a short-term sign ifican t increase in clinical comp lete response
rate and a long-term significant locoregional relapse-free survival in premenopausal patients. So, it seemed worth
comparing this regimen with CMF as adjuvant chemotherapy in moderate risk operable breast carcinoma. Methods:
Four hundred and eighty-nine patients with stage I or II moderate risk breast carcinoma were randomized to receive
CMF or CMFEV regimen for 6 cycles after surgery. Main end points were overall survival (OS), invasive disease-free
survival (IDFS) and recurrence-free interval (RFI), as estimated by Kaplan-Meier analyses and log-rank tests. Results:
At a median observation time of 7.3 years (range 5.4 months-10.3 years), no significant differences in OS and IDFS
were observed between the two arms. Deaths from breast carcinoma were more frequent with CMF (58.5%) than with
CMFEV regimen (41.7%) as well as recurrences from breast carcinoma (58.8% with CMF and 41.2% with CMFEV).
These differences were not statistically significant. Conclusion: CMFEV appears more effective than CMF in prevent-
ing recurrences from primary disease in patients with moderate risk stage I-II breast carcinoma. The lack of statistical
significance of th e ob served differen ces was p roba b ly du e to the limited number o f patients en rolled which rend ered the
study underpowdered.
Keywords: Breast Carcinoma, Adjuvant Chemotherapy, CMF Regimen, Epirubicin, Vincristine, Second Mali gna ncy
1. Introduction
The role of adjuvant systemic therapy in early stage re-
sectable breast carcinoma has been established in a
number of prospective randomized studies, and its sig-
nificant contribution in reducing the odds of relapse and
death has been clearly validated by the worldwide over-
view [1].
*On behalf of the Italian Oncology Group for Clinical Research (GOIRC),
Parma, Italy. The Medical Oncology Units of Palermo, Foligno, Todi and
Grosseto contributed with only a few cases.
The Milan Cancer Institute research group activated
the first studies demonstrating and confirming the long

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 343
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
term efficacy of the combination of cyclophosphamide,
methotrexate, 5-fluorouracil (CMF), which became a
classical chemotherapy regimen [2,3]. Since the first re-
port, this combination has been modified in a number of
ways, but mainly by the addition of other drugs, such
anthracycline, sometimes vincristine and, lately, taxanes
[4-8].
The CMFEV (cyclophosphamide, methotrexate, 5-
fluorouracil, epirubicin, vincristine) regimen is an inno-
vative schedule, compared to CMF, aimed at administer-
ing five partially or totally no cross-resistant cytotoxic
agents. It was first designed and tested by our Group as a
means of late intensification after CMF in metastatic
breast carcinoma [9] and then in the neoadjuvant setting
of operable breast carcinoma [10,11]. Its rotational strat-
egy is different from that of alternating or sequential
schemes as the five agents are administered at full dose
but, in order to avoid excessive toxicity and consequent
dose reductions, each cycle involves the administration
of only four drugs, always including vincristine (V) and
epirubicin (E). The regimen is organized in such a way
that only two, among the three potentially myelotoxic
drugs of the CMF regimen, are rotatively included in the
schedule (CMEV, CFEV, MFEV). The planned dosages
of C, M and F in each cycle were therefore either 100%
or 0%.
The aim of this prospective randomized study was to
compare the classical CMF regimen with the CMFEV
rotational regimen, both administered postoperatively for
6 cycles, in patients with operable breast carcinoma at a
moderate risk of relapse (1 to 3 positive axillary nodes or
axillary node negative with at least one biological or
morphological risk factor).
2. Patients and Methods
2.1. Eligibility Criteria
The main eligibility criteria were 1) histologically proven
operable breast carcinoma recently submitted to a poten-
tially curative surgery; 2) 1 to 3 axillary positive nodes or
negative axillary nodes with at least one biological risk
factor (estrogen and progesterone receptor negative
and/or high proliferative activity (Ki 67, or another pro-
liferative index, higher than 15%) and/or histological
grade 3; 3) age ≤70 years; 4) absence of previous or
concomitant contralateral breast carcinoma or of previ-
ous or concomitant different malignant neoplasm; 5) ab-
sence of distant metastases following a complete staging
including physical examination, chest X-ray, bone scan,
liver echography or computed tomography; adequate
bone marrow, kidney, liver, and heart function. Patients
with clinical stage III tumours (T3N1; or T4 any N; or
any T N2) were not eligible.
2.2. Study Design and Treatment
This was a multi-institutional study carried out by the
Medical Oncology Units of Parma, Reggio Emilia, Pe-
rugia, Piacenza, Terni, Vigevano, Fermo and Marsciano
of the Italian Oncology Group for Clinical Research
(GOIRC). The Medical Oncology Units of Palermo, Fo-
ligno, Todi and Grosseto contributed with a few cases.
The study design was approved by the ethical com-
mittees of the participating institutions and conducted in
accordance with the International Good Clinical Practice
Guidelines. All patients gave their written informed con-
sent before enrolment in the study. The patients were
centrally randomized via phone call to the coordinating
office of the GOIRC in Parma. Allocation was made
within strata defined by institution, menopausal status
(premenopausal vs. postmenopausal), tumour diameter
(T1, T2, T3), hormonal receptor status (negative or posi-
tive), axillary nodal status (positive or negative). Patients
were assigned to receive 6 cycles of CMF regimen or 6
cycles of CMFEV rotational regimen.
Doses and schedules of the CMF combination and of
the CMFEV rotational combination are reported in Table
1.
Postmenopausal patients, independently from their es-
trogen receptor (ER) and/or progesterone receptor (PgR)
status, received oral tamoxifen 20 mg per day for 5 years,
starting at the end of the chemotherapy treatment.
Premenopausal patients received this treatment with oral
tamoxifen only if they had their estrogen and/or proges-
terone receptor status positive.
Blood chemistry and liver function tests were repeated
on day 1 of each cycle, and complete blood counts were
obtained on day 1 and 8. Treatment was delayed of one
week if the white blood cell (WBC) count was lower
than 4000 and/or platelet count was lower than 120,000.
On day 1 after the one week delay, and on day 8, the
dosages of C, M, F, and E were reduced by 30% when
the WBC ranged from 3900 to 3600 and/or the platelet
count ranged from 119,000 to 100,000, and by 50%
when the WBC ranged from 3500 to 2500 and/or the
platelet count ranged from 99,000 to 70,000. No drugs
were administered when the WBC was less than 2500
and/or the platelet count was less than 70,000. No dose
reductions were planned for vincristine.
The adjuvant chemotherapy program had to begin no
later than 6 weeks from the initial surgery. Radiation
therapy, whenever indicated, had to start after the con-
clusion of chemotherapy program.
Treatment with tamoxifen (20 mg/day for 5 years),

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 344
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
Table 1. CMF and CMFEV re gimens: Doses and schedules.
CMF (standard regimen)
Cyclophosphamide: 600 mg/m2, iv short infusion day 1 and 8
Methotrexate: 40 mg/m2, iv bolus day 1 and 8
5-Fluorouracil: 600 mg/m2, iv bolus day 1 and 8
(every 4 weeks, 6 cycles)
CMFEV (rotational regimen)
CMEV combination
Cyclophosphamide: 600 mg/m2, iv short infusion, day 1 and 8
Methotrexate: 40 mg/m2, iv bolus, day 1 and 8
Epirubicin: 40 mg/m2, iv bolus, day 1 and 8
Vincristine: 1.4 mg/m2, iv bolus day 1
(every 4 weeks, cycles 1 and 4)
CFEV combination
Cyclophosphamide: 600 mg/m2, iv short infusion day 1 and 8
5-Fluorouracil: 600 mg/m2, iv bolus, day 1 and 8
Epirubicin: 40 mg/m2, iv bolus day 1 and 8
Vincristine: 1.4 mg/m2, iv bolus day 1
(every 4 weeks, cycles 2 and 5)
MFEV combination
Methotrexate: 40 mg/m2, iv bolus, day 1 and 8
5-Fluorouracil: 600 mg/m2, iv bolus, day 1 and 8
Epirubicin: 40 mg/m2, iv bolus, day 1 and 8
Vincristine: 1.4 mg/m2, iv bolus, day 1
(every 4 weeks, cycles 3 and 6)
when administered, began after the last cycle of chemo-
therapy Cardiotoxicity was monitored evaluating the left
ventricular ejection fraction by radionuclide or ultra-
sound technique at baseline and, in patients assigned to
the CMFEV arm, at the end of the third and of the sixth
cycle.
Follow-up visits took place every three months during
the first 2 years, every six months during years 3 through
5, and annually thereafter.
2.3. Toxicity Evaluation
Toxicity was evaluated according to the WHO criteria
[12], and the patients were classified on the basis of the
worst degree of treatment related side effects. .
2.4. Endpoints and Statistical Analyses
The definition of the endpoints selected for this study
follows the recommendations of the STEEP system [13].
The primary end point was overall survival (OS), as
estimated from the date of random assignment to the date
of last contact or death from any cause. Secondary end-
points were the invasive disease-free survival (IDFS) and
the recurrence-free interval (RFI). IDFS was estimated
from the date of random assignment to the date of occur-
rence of any of invasive ipsilateral breast tumor recur-
rence, locoregional recurrence, distant recurrence, inva-
sive contralateral breast cancer, second primary invasive
cancer, or death from any cause, whichever came first.
RFI was estimated from the date of random assignment to
the date of occurrence of any event related to the primary
breast tumour, i.e. ipsilateral breast, locoregional or dis-
tant recurrence, or death from breast cancer. All random-
ized patients were included in the estimations of OS, IDFS
and RFI, according to the intention-to-treat principle.
Others aims of the study were the estimates of the lo-
coregional recurrence-free survival (LRRFS) and of the
distant recurrence-free survival (DRFS).
To estimate the LRRFS, the recurrence of invasive
carcinoma in the ipsilateral breast, chest wall or skin, or in
the ipsilateral axillary, supraclavicular or internal mam-
mary lymph nodes were considered as events. To estimate
the DRFS, the first occurrence of metastasis at any distant
site was considered as event. In the RFI, LRRFS and
DRFS analyses, the patients who developed a contralat-

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 345
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
one at a time. The likelihood ratio test was used to evalu eral primary breast carcinoma or a second primary
non-breast malignancy were censored. ate the statistical significance of each interaction term. The
results of the subgroup analyses are graphically summa-
rized using the Forest plot as indicated by Cuzick [16].
OS, IDFS, RFI, LRRFS and DRFS were obtained from
Kaplan–Meier analyses [14], and the primary comparison
between the two groups was carried out using the log-rank
test.
The Pearson Chi-square test and the Fisher’s exact test
were used to compare the distribution of patient charac-
teristics and toxicities in the two treatment arms.
Cox’s model [15] was used for multivariate analyses to
assess the independent prognostic role of each prognostic
factor, while adjusting for the effect of the other factors.
The variables included in the models as covariates were:
treatment assigned (CMF or CMFEV), patient age (≤40
years, 41 - 50 years, 51 - 60 years or >60 years), meno-
pausal status (pre or post), clinical T (T1 or >T1), grading
(G1/G2 or G3), lymph node status (positive or negative)
and hormonal receptor status (ER–/PgR–, ER+/PgR+ or
either one receptor positive). Hazard ratios (HRs) for each
variable were obtained by exponentiating the coefficients
estimated by the Cox models. Modifications of the relative
effect of CMFEV as compared to CMF across the strata of
each covariate were assessed by introducing the appropri-
ate interaction terms in the model. These covariates by
treatment interaction terms were introduced in the model
All statistical tests were two-sided and were carried out
using the SPSS package (version 13.0 for Windows).
Significance was accepted for P values <0.05.
3. Results
3.1. Patients Characteristics
Between October 1994 and April 2000, 489 patients were
randomized to receive CMF (n = 244, 49.9%) or CMFEV
(n = 245, 50.1%). One patient, assigned to the CMF arm,
was not eligible due to the presence of 7 positive axillary
nodes.
Table 2 summarizes the main characteristics of the
randomized patients. The median age was 54 years both
in CMF arm (range 31 to 70) and in CMFEV arm (range
Table 2. Patient characteristics.
Total, n (%) CMF, n (%) CMFEV, n (%) p #
Age
Median (y)
≤40 y
>40 e ≤50 y
>50 e ≤60 y
>60 y
54 (11.1)
135 (27.6)
187 (38.2)
113 (23.1)
54 (31 - 70)
26 (10.6)
67 (27.4)
96 (39.4)
55 (22.6)
54 (29 - 70)
28 (11.4)
68 (27.8)
91 (37.1)
58 (23.7)
0.961
Menopausal status
Premenop
Postmenop ≤60 y
Postmenop > 60 y
209 (42.7)
167 (34.2)
113 (23.1)
103 (42.2)
86 (35.2)
55 (22.6)
106 (43.3)
81 (33.1)
58 (23.6)
0.874
Tumor diameter
≤2 cm.
>2 cm.
260 (53.2)
229 (46.8)
132 (54.1)
112 (45.9)
128 (52.2)
117 (47.8)
0.681
Nodal status
Negative
Positive
203 (41.5)
286 (58.5)
104 (42.6)
140 (57.4)
99 (40.4)
146 (59.6)
0.619
Histological grade
G1 + G2
G3
Unknown
216 (44.2)
207 (42.3)
66 (13.5)
113 (46.3)
98 (40.2)
33 (13.5)
103 (42)
109 (44.5)
33 (13.5)
0.339
Receptor status
ER+/PgR+
ER–/PgR–
Either one positive
249 (50.9)
139 (28.4)
101 (20.7)
127 (52.1)
71 (29.1)
46 (18.8)
122 (49.7)
68 (27.8)
55 (22.5)
0.741
Proliferative
activity (Ki67)
<15%
>15%
98 (20.0)
391 (80.0)
41 (16.8)
203 (83.2)
57 (23.2)
188 (76.8)
0.074
Type of surgery
Breast-conserving
Mastectomy
296 (60.5)
193 (39.5)
154 (63.1)
90 (36.9)
142 (57.9)
103 (42.1)
0.243
Two hundred forty-four (49.9%) patients in the CMF arm and 245 (50.1%) patients in the CMFEV arm. #Pearson Chi –Square Test for hetero-
geneity. ER, estrogen receptor; PgR, progesterone receptor.

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 346
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
29 to 70). Two hundred and nine (42.7%) patients were
premenopausal; overall, 280 (57.3%) patients were
postmenopausal; among those, 167 (34.2%) were <60
years and 113 (23.1%) were >60 years. Two hundred and
sixty (53.2%) patients had tumour diameter <2 cm and
229 (46.8%) >2 cm. Two hundred and three (41.5%)
patients were node-negative and 286 (58.5%) patients
were node-positive. Histological grade G1 or G2 were
found in 216 (44.2%) patients and G3 in 207 (42.3%)
patients. ER and PgR were both positive in 249 (50.9%)
patients and both negative in 139 (28.4%) patients. Either
one of the receptors was positive in 101 (20.7%) of the
cases. Tumour proliferative activity was low in 98
(20.0%) patients and moderate/high in 391 (80.0%) pa-
tients. Two hundred ninety-six (60.5%) patients received
breast conserving surgery, and the remaining 193 (39.5%)
underwent mastectomy. No remarkable differences in the
patient characteristic distribution between the two study
arms were seen (all p > 0.05).
3.2. Survival and Events
The distribution of all events (death, recurrence or new
malignancy) between the two arms is summarized in
Table 3.
Sample size estimates, at the time of the design of this
trial, were based on unrealistic and outdated projections
about survival and effects of the experimental treatment
(e.g. 5-year OS = 65%). Posterior power estimates, based
on the number of events actually observed, indicate that
the study, with 120 relapses, had a power of 89% to de-
tect HR’s of 0.6 for IDFS. With 63 deaths, the study had
standard power (80%) to detect only risk reductions in
excess of 50%.
3.2.1. Overall Survival
The cut-off date for follow-up was July 31, 2006. The
median observation time from random assignment to
death or censoring was 7.31 years (range: 5.4 months
-10.3 years). At the end of the observation period, 426
patients (87.1 %) were alive, with a median follow-up of
7.68 years (7.73 years for the CMF arm and 7.64 years
for the CMFEV arm). Among those alive patients, 36
(85.4%) were disease-free and 62 (14.6%) were not.
Overall, 63 deaths occurred, 33 (52.4%) in the CMF arm
and 30 (47.6%) in the CMFEV arm. Among these deaths,
48 (76.2%) were a consequence of the primary breast
tumor and 15 (23.8%) were due to other causes. No sig-
nificant difference in OS was seen between the two arms
(Figure 1, log rank p = 0.687). Cumulative OS at 5 years
was 93.0% (95% CI 91.2 - 94.4) in the CMFEV arm and
92.6% (95% CI 90.8 - 94.1) in the CMF arm. At 10 years,
these values decreased to 80.5% (95% CI 77.9 - 82.4) in
the CMFEV arm and 82.3 (95% CI 79.8 - 84.5) in the
CMF arm.
3.2.2. Invasive Disease-Free Survival
Overall, 120 events were observed, 61(50.8%) in the
CMF arm and 59 (49.2%) in the CMFEV arm. Among
these events, 1 case of death from primary breast cancer
without relapse was observed. The 120 events thus in-
cluded 90 locoregional or distant recurrences and 30
cases of second malingnacy or contralateral breast cancer.
IDFS, as estimated for all these events, was not statisti-
cally different between the two arms (log rank p = 0.892,
not shown).
3.2.3. Recurrence-Free Interval, Locoregional
Recurrence-Free Survival and Distant
Recurrence-Free Survival
When considering the RFI, which evaluates only the
events related to the primary breast tumor, we observed
that the rate of recurrence was higher in the CMF arm
than in the CMFEV arm. Of the 90 recurrences, 53
(58.9%) occurred in the CMF arm and only 37 (41.1%)
Table 3. Events according to treatment.
EVENT CMF arm n (%) CMFEV arm n (%)TOTAL n (%)
Death
from breast cancer
from other cause
total
28 (58.5)
5 (33.3)
33 (52.4)
20 (41.7)
10 (66.7)
30 (47.6)
48 (76.2)
15 (23.8)
63 (100.0)
Recurrence
locoregional
distant
¶
total
15 (62.5)
38 (57.5)
53 (58.9)
9 (37.5)
28 (42.4)
37 (41.1)
24 (26.7)
66 (73.3)
90 (100.0)
Second malignancies
contralateral breast
endometrium
leukemia
others
§
total
2 (33.3)
3 (50.0)
0 -
3 (20.0)
8 (26.7)
4 (66.7)
3 (50.0)
3 (100.0)
12 (80.0)
22 (73.3)
6 (20.0)
6 (20.0)
3 (10.0)
15 (50.0)
30 (100.0)
¶ including the one patient who died without revealed recurrence; § including ovary

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 347
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
OS 95%CIOSYears
5
10
N
215
12 93.0 80.5 (91.2-94.4) (77.9-82.4)
CMF EV
N
216
10 92.6 82.3
95%CI
(90.8-94.1)
(79.8-84.5)
CMF
P= 0.687
CMFEV
CMF
Overall Survival
Years
Proportion Surviving
Figure 1. Kaplan–Meier analysis of overall survival. CMF, cyclophosphamide, methotrexate and 5’ fluorouracil; CMFEV,
CMF, epirubicin and vincristine; N, number of patients at risk; OS, overall survival with 95% confidence interval (CI) in
parentheses. P value from log-rank test (two-sided) = 0.687.
in the CMFEV arm. This beneficial effect, although not
statistically significant, of CMFEV over CMF is shown
in Figure 2 At 10 years, recurrence-free was 80.6%
(95% CI 72.1 - 87.5) in the CMFEV arm as compared to
73.5% (95% CI 63.5 - 80.8) in the CMF arm (log rank p
= 0.099).
Of these 90 events, 24 were locoregional recurrences
and 66 were distant recurrences. Among the 24 locore-
gional events, 15 (62.5%) were in the CMF arm and 9
(37.5%) in the CMFEV arm. Cumulative 10-year LRRFS
was not statistically different between the two arms, be-
ing 90.4% (95% CI 88.4 - 92.1) for the CMF arm and
94.0% (95% CI 92.3 - 95.3) for the CMFEV arm (log
rank p = 0.472; not shown). Among the 66 distant re-
lapses, 38 (57.5%) were in the CMF arm and 28 (42.4%)
in the CMFEV arm. Cumulative 10-year DRFS was
similar in the 2 arms, being 78.7% (95% CI 76.1 - 81.1)
for the CMF arm and 83.5% (95% CI 81.1 - 85.7) for the
CMFEV arm (log rank p = 0.231; not shown).
3.2.4. Second Malignan cies
Overall, 30 cases of second malignancies were observed.
These events were significantly more frequent in the
CMFEV arm than in the CMF arm (Odd ratio = 3.53,
95% CI 1.38 - 9.26, p = 0.003). In the CMF arm, second
malignancies occurred in 8 patients (2 contralateral
breast cancers and 6 non-breast malignancies); in the
CMFEV arm, second malignancies occurred in 22 pa-
tients (4 contralateral breast cancers and 18 non-breast
malignancies). In the CMF arm, new primary non-breast
malignancies were endometrial cancer (3 patients), ovary,
lung or kidney cancer (1 patient for each site). In the
CMFEV arm, the second non-breast malignancies were
melanoma (1 patient), endometrial cancer (3 patients),
ovary cancer (3 patients), colorectal cancer (2 patients),
thyroid, liver, lung, stomach or pancreas cancer (1 patient
for each site), glioblastoma (1 patient) and leukemia (3
patients).
3.3. Multivariate and Subgroup Analyses
In multivariate analyses, nodal status, tumor size and
hormonal receptor status were independently associated
with OS, IDFS and RFI (not shown) After adjustment for
these three factors as well as for patient age, histological
grade and menopausal status, we did not find any sig-
nificant difference in the hazard of death between the
CMFEV arm and the CMF arm in the overall population
(HR = 0.80, 95% CI 0.48 - 1.35, p = 0.411, Figure 3). A
similar lack of treatment effect on IDFS was observed
(HR = 0.91, 95% CI 0.62 - 1.34, p = 0.645, not shown).
By contrast, the hazard of recurrence from the primary
breast tumor was lower, with a borderline statistical sig-
nificance, in the CMFEV arm as compared to the CMF
arm (HR = 0.67, 95% CI 0.43 - 1.04, p = 0.073, Figure
4). Subgroup analyses of OS comparing the CMFEV arm
versus the CMF arm within strata formed by each prog-
nostic factor showed evidence of interaction between the
type of adjuvant treatment and the receptor status (Fig-

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 348
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
121086420
100
80
60
40
20
0
CMFEV
CMF
Recurrence
-
Free Interval
P= 0.099
RFI 95%CIRFIYears
5
10
N
86.5
80.6 (77.7-91.6) (72.1-87.5)
CMFEV
N
203
28 85.0 73.5
95%CI
(76.6-90.8)
(63.5-80.8)
CMF
187
26
Proport ion Recurren ce-F ree
Figure 2. Kaplan–Meier analysis of recurrence-free interval CMF, cyclophosphamide, methotrexate and 5’ fluorouracil;
CMFEV, CMF, epirubicin and vincristine; N, number of patients at risk; OS, overall survival; EFS, recurrence-free interval,
with 95% confidence interval (CI) in parentheses. P value from log-rank test (two-sided) = 0.099.
C lin ic al T
Ag e group
Tu mor grad e
Me nopause
Nodal Status
Receptor Status
Overall (C MFEV vs CMF)
<= 40 y .
41 –50 y.
51 –60 y.
>60 y.
T1
> T1
G1/G2
G3
Pre
Post
Neg.
Pos.
ER -/PgR -
Either one +
ER +/PgR +
0.655
0.784
0.526
0.345
0.138
0.026
1.03 (0.44 –2.40)
0.74 ( 0.32 – 1.71)
1.13 (0.33 –3.86)
0.71 (0.21 –2.40)
0.36 (0.13 –0.97)
1.22 (0.42 –3.58)
0.94 (0.45 –1.94)
0.70 (0.33 –1.51)
1.04 (0.47 -2.32)
0.53 (0.25 –1.09)
0.94 (0.43 –2.07)
0.69 (0.37 –1.41)
1.32 (0.47 –3.72)
0.62 (0.33 -1.16)
0.60 (0.19 –1.88)
0.80 (0.48 –1.35)0.411
#
¶
#
#
#
#
#
CMFE V be tte rCMF be tte r
H az ard ratio (CM FEV vs CMF) and 95% CI
0.3 14
OVE RAL L SURVIVAL
Figure 3. Forest plot of subgroup analysis of OS comparing the CMFEV arm versus the CMF arm within strata formed by
each prognostic factor. Hazard ratios and 95% confidence intervals (CI) from a Cox Model in which all covariates signifi-
cantly contributing to the likehood of the model in the entire dataset were used. Interaction terms assessing the heterogeneity
of the effect of treatment regimens across strata for each covariate were introduced in the model one at a time. P values are
from likehood ratio tests. All statistical tests were two-sided. The plain line shows no effect point and the dotted line shows
overall treatment effect for the entire dataset. #test for interaction; ¶overall comparison of CMFEV arm versus CMF arm
adjusted for all prognostic factors.

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 349
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
CMF E V be tt e r
0.1 51
CMF be tte r
C li n ica l T
A g e gr oup
Tu mor grade
Menopaus e
Nodal Status
Rece ptor Status
<= 40 y .
41 –50 y.
51 –60 y.
>60 y.
T1
> T1
G1/G2
G3
Pre
Post
Neg.
Pos.
ER -/PgR -
Either one +
ER +/PgR +
0.233
0.203
0.149
0.460
0.57 (0.15 –2.25)
0.65 (0.28 –1.53)
0.57 (0.26 –1.21)
0.88 (0.36 – 2.17)
0.64 (0.33 –1.25)
0.71 (0.40 –1.29)
0.88 (0.47 –1.63)
0.47 (0.25 –0.91)
0.60 (0.30 –1.19)
0.74 (0.41 –1.32)
1.06 (0.22 –5.01)
0.77 (0.31 –1.91)
1.03 (0.47 –2.24)
0.55 (0.22 –1.38)
0.48 ( 0.24 – 0.98)
0.67 (0.43 –1.04)0.073
0.002
#
¶
0.338
#
#
#
#
#
Hazard ratio
(
CMFEV vs CMF
)
and 95% CI
Overall (CMFEV vs. CMF)
RECURRENCE-FREE INTERVAL
Figure 4. Forest plot of subgroup analysis of RFS comparing the CMFEV arm versus the CMF arm within strata formed by
each prognostic factor. Hazard ratios and 95% confide nce intervals (CI) from a Cox Model in which all covariates significantly
contributing to the likehood of the model in the entire dataset were used. Interaction terms assessing the heterogeneity of the
effect of treatment regimens across strata for each covariate were introduced in the model one at a time. P values are from like-
hood ratio tests. All statistical tests were two-sided. The plain line shows no effect point and the dotted line shows overall treat-
ment effect for the entire dataset. #test for interaction; ¶overall comparison of CMFEV arm versus CMF arm adjusted for all
prognostic factors.
Table 4. Main toxicities.
CMF arm CMFEV arm
n%n%
Hematological toxicities
Hemoglobin
G1/2 34147129
G3/4 7318 7
WBC
G1/2 1044393 38
G3/4 17744 18
Platelets
G1/2 4294
Non haematological toxicities
Nausea/vomiting 1606617973
Epigastric pain 27114016
Constipation 62156
Oral 532210944
Mucositis 71296928
Skin 104156
Kidney 2121
Liver 30122912
Heart 156177
Asthenia 27114117
Mandibular pain 00146
Peripheral nervous system 1677430

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 350
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
ure 3, p = 0.026). CMFEV regimen was more favourable
when patients presented both estrogen and progesterone
receptors positive, or either one receptor positive (HR =
0.74, 95% CI 0.32 - 1.71 and HR = 0.60, 95% CI 0.19 -
1.88, respectively). By contrast, no treatment effect was
seen when both receptors were negative (HR = 1.03, 95%
CI 0.44 - 2.40) Similar results were obtained for sub-
group analyses of RFI (Figure 4, p = 0.002; HR = 0.48,
95% CI 0.24 - 0.98 for ER+/PgR+; HR = 0.55, 95% CI
0.22 - 1.38 for either one receptor positive; HR = 1.0,
95% CI 0.47-2.24 for ER-/PgR). Regarding IDFS sub-
group analysis, no significant treatment effect with re-
spect to positive receptors was observed (not shown).
3.4. Toxicities
Table 4 reports the main toxicities according to treat-
ment. With respect to haematological toxicities, there
was a higher incidence of anaemia both grade 1 - 2 and 3
- 4 in the CMFEV arm than in the CMF arm (29% vs.
14%, p = < 0.001 and 7% vs. 3%, p = 0.02, respectively);
there were no relevant differences between the two arms
in terms of WBC and platelet toxicities.
With respect to non-haematological toxicities, in the
CMFEV arm, compared to the CMF arm, there was a
more frequent occurrence of stomatitis (44% vs. 22%; p
< 0.001), constipation (6% vs. 2%; p = 0.04), peripheral
nervous system toxicity (30% vs. 7%; p < 0.001), man-
dibular pain (6% vs. 0%).
4. Discussion
4.1. CMFEV Regimen versus CMF Regimen:
General Considerations
In the present study, our Group compared for the second
time the conventional CMF regimen with the experi-
mental CMFEV regimen in the treatment of non metas-
tatic breast carcinoma. In the first occasion, the com-
parison was developed in a neoadjuvant settings [10,11];
in the present second occasion, the comparison was de-
veloped in an adjuvant settings. Therefore such com-
parison presents, by itself, some similarities but also
some substantial differences. Firstly, these differences
are related to the therapeutic scheme utilized, which re-
fers to the different treatment settings. Moreover, in the
neoadjuvant settings, the comparison of the therapeutic
results could be done both on a short term, based on the
evaluation of objective therapeutic response [10], as well
as on a long term, based on the outcome [11]. In the ad-
juvant settings the comparison was possible only on a
long term.
In addition, in the present study, some results, which
confirmed a superiority of the CMFEV regimen over the
CMF regimen, were possibly disturbed and masked by an
unexpected increase of second malignancies observed in
the CMFEV arm.
For all these reasons, we believe that in a first part of
the discussion, it could be useful to summarize in a
comparative way the number and the proportions of the
main events (relapse and death) we observed in the pre-
sent study when administering CMF or CMFEV. In a
second part of the discussion, we will elaborate a com-
parative evaluation of the main results observed in our
first study of neoadjuvant chemotherapy and in the pre-
sent study of adjuvant chemotherapy, in terms of efficacy,
relationship with endocrine parameters, toxicities. In a
third part, we will comment on the unexpected increase
of second malignant tumors here observed with CMFEV.
In the last part of the discussion, we will compare our
results to those reported by other Authors within a com-
prehensive evaluation of the adjuvant chemotherapy of
operable breast carcinoma.
4.2. CMFEV Regimen versus CMF Regimen:
Efficacy, Relationship with Endocrine
Parameters and Toxicity
The results of the present study showed that the recur-
rence from the primary tumor was more frequent with
CMF regimen (63 events, 58.2%) than with CMFEV
regimen (38 events, 41.8%). At 10 years, RFI was 73.5%
with CMF and 80.6% with CMFEV. Similarly, deaths
due to primary tumor were more frequent in the CMF
arm (28, 58.5%) compared to the CMFEV arm (20,
41.7%), even if the OS at 10 years was similar with ei-
ther regimen (82.3% with CMF, 80.5% with CMFEV).
These differences in terms of recurrence and death,
although not statistically significant, offer evidence of a
potential major efficacy of the CMFEV compared to the
CMF regimen. This efficacy was, at least in part, more
clearly showed in our previous study on neo-adjuvant
chemotherapy where, on a short term evaluation, the rate
of clinical responses, both complete (CR) and complete
plus partial (PR), were significantly higher in the subset
of pre-menopausal patients treated with CMFEV com-
pared to those treated with CMF [10]. Similarly, on a
long-term evaluation, again in the subset of
pre-menopausal patients, the proportion of RFS tended to
be higher and the proportion of LRRFS was significantly
higher in the CMFEV arm compared to the CMF arm,
thus mirroring the short-term response results [11].
The reason of the lack of statistical significant differ-
ences observed in the present study and of only a few
statistically significant differences observed in the pre-

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 351
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
ent study.
vious study, is probably due to the fact that, in the two
studies, different parameters were considered and esti-
mated, and that the eligibility criteria, although similar,
were not identical. In any case, both studies may have
been under-powered to demonstrate a significant major
efficacy of CMFEV over CMF.
Interestingly, both the previous and the present studies
showed statistically significant differences between the
two chemotherapy regimens according to the menopausal
status of the patients or to the biological characteristics of
their tumors in terms of ER and/or PgR status. Indeed in
the first study, the superiority of CMFEV regimen, as
short term objective response, was seen only in pre-
menopausal and not in post-menopausal patients; in ad-
dition, in a multivariate analysis, a significant interaction
was confirmed between the menopausal status and the
type of treatment on the probability to obtain CR or CR
plus PR [10]. In the present study, a significant correla-
tion was observed between the estrogen and/or proges-
terone receptor status and the type of regimen; CMFEV
was more effective, in terms of OS and RFI, in patients
with positive ER and/or PgR tumors but not in patients
with both negative ER and PgR tumors. Considering
those correlations, it appears that endocrine influences
induced by the menopausal status of the patients and/or
by the ER and/or PgR status of the tumors may deter-
mine the comparative responses of CMF / CMFEV used
either as neo-adjuvant or adjuvant chemotherapy in op-
erable breast carcinoma.
As to the main toxicities observed in the present study,
the higher proportion of constipation, mandibular pain
and peripheral nervous system toxicity on the CMFEV,
compared to CMF regimen, could be due to the addition
of vincristine. In the first study, a higher proportion of
mild neurological side effects was also observed when
administering CMFEV as compared to CMF [10]. The
anaemia and the stomatitis could also be attributed to
vincristine, or to the addition of epirubicin. Overall, the
toxicities reported after the administration of CMFEV
were rather well tolerated; only in a few patients, the
treatment had to be shortened or vincristine administra-
tion had to be discontinued.
4.3. CMFEV Regimen versus CMF Regimen:
Incidence of Second Malignancies
The significantly higher incidence of second malignan-
cies with CMFEV as compared to CMF was unexpected
and deserves some consideration as we did not observed
this difference when using these regimens in a
neo-adjuvant settings [10] This discrepancy may result
either from the higher number of patients in the present
study (489 versus 211) or from differences in the number
of chemotherapy cycles delivered (4 cycles as
neo-adjuvant therapy versus 6 cycles as adjuvant ther-
apy). Moreover, epirubicin and vincristine were admin-
istered in 3 and 4 cycles, respectively, in the previous
study [10] compared to 6 cycles in the pres
4.4. CMFEV Regimen versus CMF Regimen:
Comprehensive Evaluation of the Adjuvant
Chemotherapy of Operable Breast
Carcinoma
Two overviews had reported that, in the adjuvant che-
motherapy of operable breast cancer, anthracycline con-
taining regimens were more effective than CMF in pre-
venting recurrence and death [1,4]. This concept was
particularly stressed by the former study that reported
highly statistically significant differences in favour of the
anthracycline-containing regimens as compared to CMF
[1]. It has to be noted that this overview focused on the
anthracycline-containing regimens FAC (fluorouracil,
adriamycin/doxorubicin, cyclophosphamide) and FEC
(fluorouracil, epirubicin, cyclophosphamide) while the
anthracycline-containing regimens AC (adriamycin/,
cyclophosphamide) and EC (epirubicin, cyclophos-
phamide) were not considered [1]. According to the clas-
sification adopted by us in 2003 [5], the administration of
anthracycline in the AC and EC regimens was substan-
tially “substitutive” as only one of the three CMF agents
(cyclophosphamide) was maintained, while in the FAC
and FEC regimens the administration of anthracycline
was substantially “additive” as two of the three agents of
CMF (cyclophosphamide and 5-fluorouracil) were main-
tained. On the other hand, considering the results of single
studies where more than 2000 patients were enrolled,
comparisons of CMF with the AC regimen did not pro-
vide evidences favouring AC [17,18], while comparisons
of CMF with the FAC regimen or with the FEC regimen
reported results significantly favouring these two anthra-
cycline-containing combinations [19,20].
In this light, our CMFEV experimental regimen
clearly belongs to the “additive” type of combinations
but it exceeds FAC or FEC combinations as it involves
the conservation of not only two, but of all three CMF
agents, although in a rotational strategy. This considera-
tion together with the results we obtained with the
CMFEV regimen, both in the first study and in the pre-
sent one, should encourage the conduction of further
studies but the unexpected increase of second malignan-
cies here observed upon CMFEV treatment suggests
some cautions. In addition, at the present time, any dis-
cussion about the adjuvant chemotherapy of breast car-

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with 352
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
cinoma should not exclude the consideration of the addi-
tion of taxanes to anthracyclines although results re-
ported so far in this line do not sufficiently encourage
this approach [21,22].
4.5. Conclusions
This study provides preliminary evidence suggesting a
potentially higher efficacy of the CMFEV regimen as
compared to the standard CMF regimen. This result is in
line with the overall evidence demonstrating an increased
efficacy of anthracycline-containing adjuvant regimens
in breast cancer [1], but also lends to support the ration-
ale of using a 5-drugs, rotational regimen. However, be-
fore this rationale can be further explored in larger, more
focused trials, the increased incidence of cancer at other
sites here observed in the CMFEV arm, needs to be
thoroughly considered.
5. Acknowledgements
This work was supported by a grant from Associazione
Italiana per la Ricerca sul Cancro (AIRC).
The authors would like to thank dr. Paolo Bruzzi for
his statistical help in the preparation of the manuscript.
REFERENCES
[1] Early Breast Cancer Trialists’ Collaborative Group, “Ef-
fect of Chemotherapy and Hormonal Therapy for Early
Breast Cancer on Recurrence and 15-Year Survival: An
Overview of the Randomised Trials,” Lancet, Vol. 365,
No. 9472, 2005, pp. 1687-1717.
doi:10.1016/S0140-6736(05)66544-0
[2] G. Bonadonna, E. Brusamolino, P. Valagussa, A. Rossi, L.
Brugnatelli, C. Brambilla, M. De Lena, G. Tancini, E.
Bajetta, R. Musumeci and U. Veronesi, “Combination
Chemotherapy as an Adjuvant Treatment in Operable
Breast Cancer,” The New England Journal of Medicine,
Vol. 294, No. 8, 1976, pp. 405-410.
doi:10.1056/NEJM197602192940801
[3] G. Bonadonna, A. Moliterni, M. Zambetti, M. G.
Dandone, S. Pilotti, L. Gianni and P. Valagussa, “30
Years’ Follow up of Randomised Studies of Adjuvant
CMF in Operable Breast Cancer: Cohort Study,” British
Medicine Journal, Vol. 330, No. 7485, 2005, pp. 205-206.
doi:10.1136/bmj.38314.622095.8F
[4] Early Breast Cancer Trialists Collaborative Group, “Poly-
chemotherapy for Early Breast Cancer: An Overview of
Randomised Trials,” Lancet, Vol. 352, No. 9132, 1998,
No. 9132, pp. 930-942.
doi:10.1016/S0140-6736(98)03301-7
[5] G. Cocconi, “Adjuvant Chemotherapy in Early Breast
Cancer: Optimal and Suboptimal Anthracycline Containig
Regimens,” Breast Cancer Research and Treatment, Vol.
80, No. 3, 2003, pp. 313-320.
doi:10.1023/A:1024955408785
[6] R. B Weiss, D. C.Tormey, F. Holland, V. E. Weinberg, G.
Lesnick, M. Perloff, G. Falkson and O. J. Glidewell, “A
Randomised Trial of Postoperative Five-versus Three
Drug Chemotherapy after Mastectomy; A Cancer and
Leukemia Group B (CALGB) Study,” Recent Results
Cancer Research, Vol. 80, No. 1, 1982, pp. 170-176.
[7] D. C. Tormey, V. E. Weinberg, J. F. Holland, R. B. Weiss,
O. J. Glidewell, M. Perloff, G. Falson, H. C. Falkson, P. H.
Henry and L. A. Leone, “A Randomised Trial of Five and
Three Drug Chemotherapy and Chemo-Immunotherapy in
Women with Operable Node Positive Breast Cancer. A
CALGB Study,” Journal of Clinical Oncology, Vol. 1, No.
2, 1983, pp. 138-145.
[8] L. Gianni, J. Baselga, W, Elemann, V. G. Porta, V. Semi-
glazov, A. Lluch, M. Zambetti, D. Sabadell, G. Raab, A. L.
Cussac, A. Bozhok, A. Martinez-Agullo, M. Greco, M.
Byakhov, J. J. Lopez, M. Mansutti, P. Valagussa and G.
Bonadonna, “Phase III Trial Evaluating the Addition of
Paclitaxel to Doxorubicin Followed by
Cyclo-phosphamide, Methotrexate, and Fluorouracil, as
Adjuvant or Primary Systemic Therapy European Coop-
erative Trial in Operable Breast Cancer,” Journal of
Clinical Oncology, Vol. 27, No. 15, 2009, pp. 2474-2481.
doi:10.1200/JCO.2008.19.2567
[9] G. Cocconi, G. Bisagni, M. Bacchi, F. Buzzi, R. Canaletti,
A. Carpi, G. Ceci, A. Colozza, V. De Lisi and R. Lottici
“A Comparison of Continuation versus Late Intensifica-
tion Followed by Discontinuation of Chemotherapy in
Advanced Breast Cancer. A Prospective Randomized
Trial of the Italian Oncology Group for Clinical Research
(GOIRC),” Annals of Oncology, Vol. 1, No. 1, 1990, pp.
36-44.
[10] G. Cocconi, B. Di Blasio, C. Boni, G. Bisagniu, G. Ceci,
E. Rondini, M. Bella, F. Leonardi, L. Savoldi, R. Camisa
and P. Bruzzi, “Randomized Trial Comparing Cyclo-
phosphamide, Methotrexate, and 5-Fluorouracil (CMF)
with Rotational CMF, Epirubicin and Vincristine as Pri-
mary Chemotherapy in Operable Breast Cancer,” Cancer,
Vol. 95, No. 2, 2002, pp. 228-235.
doi:10.1002/cncr.10678
[11] G. Cocconi, B. Di Blasio, C. Boni, G. Bisagni, G. Ceci, E.
Rondini, M. Bella, F. Leonardi, L. Savoldi, C. Vallisneri,
R. Camisa and P. Bruzzi, “Primary Chemotherapy in Op-
erable Breast Carcinoma Comparing CMF (Cyclophos-
phamide, Methotreaxate, 5-Fluorouracil) with an Anthra-
cycline Containing Regimen: Short-Term Responses
Translated into Long-Term Outcomes,” Annals of On-
cology, Vol. 16, No. 9, 2005, pp. 1469-1476.
doi:10.1093/annonc/mdi278
[12] A. B. Miller, R. Hoogstraten, M. Staquet and A. Winkler,
“Reporting Results of Cancer Treatment,” Cancer, Vol.
47, No. 1, 1981, pp. 207-214.
doi:10.1002/1097-0142(19810101)47:1<207::AID-CNC
R2820470134>3.0.CO;2-6
[13] C. A. Hudis, W. E. Barlow, J. P. Costantino, R. J. Gray,

Randomized Trial Comparing Cyclophosphamide, Methotrexate, and 5-Fluorouracil (CMF) Regimen with
Rotational CMFEV Regimen (E = Epirubicin, V = Vincristine) as Adjuvant Chemotherapy in
Moderate Risk Operable Breast Carcinoma
353
K. I. Pritchard, J. W. Chapman, J. A. Sparano, S. Huns-
berger, R. A. Enos, R. D. Gelber and J. O. Zujewski,
“Proposal for Standardized Definitions for Efficacy End
Points in Adjuvant Breast Cancer Trials: The STEEP
System,” Journal of Clinical Oncology, Vol. 25, No. 15,
2007, pp. 2127-2132. doi:10.1200/JCO.2006.10.3523
[14] E. L. Kaplan and P. Meier, “Nonparametric Estimation
from Incomplete Observations,” Journal of the American
Statistical Association, Vol. 53, No. 282, 1958, pp.
457-481. doi:10.2307/2281868
[15] D. R. Cox, “The Regression Analysis of Binary Se-
quences,” Journal of the Royal Statistical Society, Vol. 20,
No. 2, 1958, pp. 215-242.
[16] J. Cuzick, “Forest Plots and the Interpretation of Sub-
groups,” Lancet, Vol. 365, No. 9467, 2005, p. 1308.
doi:10.1016/S0140-6736(05)61026-4
[17] B. Fisher, A. M. Brown, N. V. Dimitrov, R. Poisson, C.
Redmond, R. G. Margolese, D. Bowman, N. Wolmark, D.
L. Wickerham, C. G. Kardinal, H. Shibata, A. H. G.
Paterson, C. M. Sutherland, N. J. Robert, P. J. Ager. L.
Levy, J. Walter, T. Wozniak, E. R. Fisher and M. Deutsch,
“Two Months of Doxorubicin-Cyclophosphamide with
and without Reinduction Therapy Compared with 6
Months of Cyclophosphamide, Methotrexate, and Fluo-
rouracil in Positive-Node Breast Cancer Patients with
Tamoxifen Non-responsive Tumors: Results from the Na-
tional Surgical Adjuvant Breast and Bowel Project B-15,”
Journal of Clinical Oncology, Vol. 8, No. 9, 1990, pp.
1483-1496.
[18] B. Fisher, S. Anderson, Chiu-Tan, N. Wolmark, D. L.
Wickerham, E. R. Fisher, N. V. Dimitrov, J. N. Atkins, N.
Abramson, S. Merajver, E. H. Romon, C. G. Kardinakl, H.
R. Shibata, R. G. Margolese, E. H. Romond, C. G.
Kardinal, H. R. Shibata, R. G. Margolese and W. B.
Farrar, “Tamoxifen and Chemotherapy for Axillary
Node-Negative, Estrogen Receptor-Negative Breast Can-
cer: Findings from the National Surgical Adjuvant Breast
and Bowel Project B-23,” Journal of Clinical Oncology,
Vol. 19, No. 4, 2001, pp. 931-942.
[19] M. Martin, A. Villar, A. Sole-Calvo, R. Gonzalez, B.
Massuti, J. Lizon, C. Camps, A. Carrato, A. Casado, M. T.
Candel, J. Albanell, J. Aranda, B. Munarriz, J. Campbell
and E. Diaz-Rubio, “Doxorubicin in Combination with
Fluorouracil and Cyclophosphamide (i.v. FAC Regimen
day 1, 21) versus Methotrexate in Combination with
Fluorouracil and Cyclophosphamide (i.v. CMF Regimen,
day 1, 21) as Adjuvant Chemotherapy for Operable
Breast Cancer: A Study of the GEICAM Group,” Annals
of Oncology, Vol. 14, No. 6, pp. 833-842.
doi:10.1093/annonc/mdg260
[20] H. T. Mouridsen, J. Andersen, M. Andersson, P. Dom-
bernowsky, B. Ejilertsen, C, Rose, P. G. Sorensen, E.
Sandberg, K. W. Andersen, M. B. Jensen, N. O. Bengston,
J. Bergh and B. Nordenskjold, “Adjuvant Anthracycline
in Breast Cancer Improves Outcome in Premenopausal
Patients Following Substitution of Methotrexate in the
CMF Combination with Epirubicin,” Proceed ASCO, Vol.
18, No. 68, 1999, p. 19.
[21] P. Francis, J. Crown, A. Di Leo, M. Buyse, A. Balil, M.
Andersson, B. Nordenskjold, I. Lang, R. Jakesz, D. Vo-
robiof, J. Gutierrez, G. van Hazel, S. Dolci, S. Jamin, B.
Bendahmane, R. D. Gelber, M. Castiglione-Gerstsch and
M. Piccart-Gebhart, “Adjuvant Chemotherapy with Se-
quential or Concurrent Anthracycline and Docetaxel:
Breast International Group 02-98 Randomized Trial,”
Journal of National Cancer Institution, Vol. 100, No. 2,
2008, pp. 121-133. doi:10.1093/jnci/djm287
[22] J. M. Albert, A. U. Budzar, R. Guzman, P. K. Allen, E. A.
Strom, G. H. Perkins, W. A. Woodward, K. E. Hoffman,
W. Tereffe, K. K. Hunt, T. A. Buchholz and G. L. Oh,
“Prospective Randomized Trial of 5-Fluorouracil, Doxo-
rubicin, and Cyclophosphamide (FAC) versus Paclitaxel
and FAC (TFAC) in Patients with Operable Breast Can-
cer: Impact of Taxane Chemotherapy on Locoregional
Control,” Breast Cancer Research and Treatment, Vol.
128, No. 2, 2011, pp. 421-427.