Background: Breast cancer is a significant cause of mortality
and morbidity worldwide, although death rates in the United States and
some other countries are beginning to fall.
Methods: Several sources of information in 1998, including
publications and presentations at the 1998 meeting of the American Society
of Clinical Oncology, are pertinent to contemporary breast cancer care.
Results: It is now possible to prescribe hormonal therapy
that will reduce the incidence of breast cancer. Methods are available
to reduce the morbidity from axillary node dissection, and improvements
in adjuvant therapy and management of metastatic breast cancer are now
at hand.
Conclusions: The information presented provides a broad-based
platform for new standards of care for breast cancer that will serve as
a sound base for further progress in this important disease.
Introduction
Breast cancer continues to be a worldwide public health
problem, and the many patients who develop the disease carry a heavy physical
and emotional burden. This year, 180,300 individuals in the United States
are expected to develop breast cancer, and 43,900 will die of its effects.
1
Some good news about the disease is already at hand. As a result of screening
programs, many institutions report that approximately one third of new
breast cancers are the highly curable
in situ lesions,
2
and mortality rates from breast cancer are now trending down.
3
During the last few months, new information about
several aspects of breast cancer has become available, including the long-awaited
publication of the meta-analysis concerning tamoxifen adjuvant therapy.4
At the recent annual meeting of the American Society of Clinical Oncology
(ASCO), several presentations highlighted significant advances pertinent
to contemporary breast cancer care. This article summarizes some of those
advances that relate to selected areas of clinical concern. Clearly, many
treatment approaches are still not ideal and controversies remain, but
the progress being made in understanding and managing breast cancer is
tangible and real.
Prevention
In the past, although it was recognized that oophorectomy
appreciably reduced breast cancer risk,
5 the only known approach
to prevent breast cancer was prophylactic mastectomy, a procedure most
commonly performed by plastic surgeons when the patient’s risk of developing
a future contralateral breast cancer was deemed to be high. Now, several
studies indicate that breast cancer incidence can be lowered by a variety
of hormonal approaches.
Tamoxifen
It has been evident for several years that the incidence
of new breast cancers in the contralateral breast of patients with breast
cancer who received tamoxifen was diminished. The latest meta-analysis
of adjuvant tamoxifen treatment,4 in fact, reports that women
who took tamoxifen for approximately five years had a 47% reduction in
the risk for a contralateral breast cancer. These data stimulated trials
of tamoxifen in women without breast cancer but who were considered to
be at high risk for developing the disease.
The recently unblinded US Breast Cancer Prevention
Trial led by the National Surgical Adjuvant Breast and Bowel Project (NSABP)
included more than 13,000 women who were at least 60 years of age, had
a history of lobular carcinoma in situ, or were between 35 and 59
years of age and also had a five-year risk of developing breast cancer
that, as judged by the Gail model,6 was at least equivalent
to that of the average 60-year-old woman. The distribution of selected
demographic characteristics is summarized in Table 1.
Table 1. --
Distribution of 12,542 Breast Cancer Prevention Trial Participants by Selected
Demographic Characteristics |
| Age of Participants |
First Degree Relatives With Breast Cancer |
Relative Risk (RR) of Breast Cancer |
| Years |
% |
# Relatives |
% |
RR |
% |
35-39 |
2.7 |
0 |
23.1 |
1.0 |
0.1 |
40-44 |
12.1 |
1 |
57.6 |
1.01-2.0 |
7.6 |
45-49 |
24.7 |
2 |
16.4 |
2.01-3.0 |
15.9 |
50-54 |
18.4 |
3 |
2.5 |
3.01-5.0 |
37.5 |
55-59 |
12.2 |
>=4 |
0.5 |
5.01-10.0 |
28.9 |
60-64 |
14.1 |
|
>10.0 |
10.0 |
65-69 |
9.9 |
|
>=70 |
5.9 |
The average relative risk of the trial participants
according to age is summarized in Table 2, and the distribution according
to the level of five-year risk of breast cancer is summarized in Table
3. This double-blind study allocated the participants to receive five years
of treatment with 20 mg of tamoxifen daily or a placebo.
Table 2. --
Average Relative Risk of 12,542 Breast Cancer Prevention Trial
Participants According to Age |
| Age (Yrs) |
Average Relative Risk |
35-39 |
9.6 |
40-44 |
7.6 |
45-49 |
6.4 |
50-54 |
5.3 |
55-59 |
5.1 |
60-64 |
3.9 |
65-69 |
3.7 |
>=70 |
3.9 |
| |
|
All Ages |
5.5 |
Table 3. --
Distribution of 12,542 Breast Cancer Prevention Trial Participants
According to Level of Five-Year Risk of Breast Cancer |
5-Year Risk of Breast Cancer (%) |
% |
<=2 |
24.7 |
2.1-3.0 |
31.6 |
3.1-4.0 |
18.4 |
4.1-5.0 |
8.5 |
>=5.1 |
16.8 |
At the 1998 ASCO plenary session, it was reported
that the mean follow-up time of study subjects in the prevention trial
was 3.6 years and more than 57% had completed four years of treatment.
One hundred fifty-four patients developed invasive breast cancers in the
placebo group vs 85 in the tamoxifen-treated group. This 45% reduction
in incidence was highly significant (P=0.00001). Benefit was evident
in all age groups but was greatest in those over 60 years. More estrogen-receptor
(ER)-positive carcinomas occurred in the placebo group, but the size of
detected tumors was equivalent, with two thirds being less than 2 cm in
diameter. The incidence of noninvasive tumors was also reduced in the tamoxifen-treated
group. Thirty-three endometrial cancers, all stage 1, developed in the
tamoxifen group vs 14 in the placebo group. There was no difference in
the incidence of gastrointestinal, liver, or ovarian cancers in the two
groups. Seventeen vascular events were noted with tamoxifen, and six were
encountered with placebo.
Raloxifene
The selective ER modulator raloxifene has estrogenic
effects on bone and lipids, but it has estrogen antagonist effects on the
breast and uterus.7 Again, at the 1998 ASCO plenary session,8
Dr Cummings described the "MORE" trial, a randomized, double-blind trial
of raloxifene vs placebo in postmenopausal women with osteoporosis under
80 years of age. This trial was designed to test the hypothesis that raloxifene
will reduce the risk of bone fractures. A total of 7,704 women were randomly
assigned to active treatment or placebo, in a ratio of 2:1, to receive
raloxifene in a dose of either 60 or 120 mg per day or a matching placebo.
After a median of 28.9 months of follow-up, 32 cases of breast cancer had
been confirmed, 11 in those patients assigned to raloxifene and 21 in those
assigned to placebo. Thus, the annual rate of developing breast cancer for raloxifene was 1.4/1000 (with no difference noted between
60 mg and 120 mg dosage levels) compared with 4/1000 for the placebo. This
gives a relative risk for raloxifene of 0.30 (P=.001). Endometrial
cancer occurred in four cases in the raloxifene group and in four cases
in the smaller placebo group. These numbers are too small to confirm a
protective effect of raloxifene on the uterus.
Inhibition of Ovulation
A third approach to breast cancer prevention was also
postulated in an educational session at the 1998 ASCO meeting (Malcolm
Pike, PhD, personal communication, 1998). Using the rationale that breast
cancer risk is related to ovulatory frequency and estrogen production,
the proposal was made to lower breast cancer incidence for premenopausal
women by reducing ovulation with a luteinizing hormone-releasing hormone
(LHRH) agonist/antagonist and by providing only small amounts of replacement
estrogen, androgens, and progestins.
Discussion
The demonstration of reduction in incidence of breast
cancer induced by two "designer" estrogens is a real finding, and raloxifene
may have an additional capacity of not increasing endometrial cancer risk.
The premature closing of the US Breast Cancer Prevention Trial is understandable,
but this action has reduced the chance that important endpoints other than
cancer incidence, such as breast cancer mortality, also could have been
made. Ongoing foreign trials that are similar in design to the US trial
may provide such data. We should note that cancers developing in subjects
who have taken tamoxifen or raloxifene will be more often ER-poor and will
likely not derive survival benefit from adjuvant hormonal interventions.
The raloxifene study, although positive in terms of reduction in breast
cancer risk, should be viewed with some caution since patients with osteoporosis
have a lower than average risk for breast cancer, and the primary focus
of the trial was on reduction of fractures rather than reduction of breast
cancer incidence. A follow-up trial enrolling 22,000 women without cancer
will open later this year and should answer questions of comparative effectiveness
of these two agents by directly comparing the cancer prevention effects
and toxicities of tamoxifen and raloxifene. Other questions, such as the
effectiveness of tamoxifen or raloxifene preventive measures in women carrying
an abnormal breast cancer susceptibility gene (eg, BRCA1), are being addressed
now.
Staging Localized Cancer
A favorite multiple choice question for medical students
asks which of several alternatives is the key prognostic factor for a patient
with clinically localized breast cancer. The correct answer, of course,
is "axillary node involvement." To quote S. Eva Singletary, MD,
9
an eminent M.D. Anderson surgeon: "With the abandonment of routine axillary
node dissection, the next hurdle for surgical therapy . . .!" Clearly,
clinical practice regarding axillary node dissection is changing.
Several issues need to be addressed when a formal
axillary dissection is not performed as part of a initial staging and treatment,
although the characteristics of the primary tumor (eg, size) are increasingly
being used for adjuvant treatment decisions. It should be remembered, however,
that there is an unacceptably high rate of local recurrence in the axilla
in clinically node-negative patients with tumors >1 cm who receive no axillary
treatment, and there is a finite incidence of axillary node metastasis
in patients with small invasive tumors and even with ductal carcinoma in
situ (DCIS).9
Sentinel node mapping is a promising new technique
that reliably predicts nodal status without the morbidity of axillary dissection.10,11
My clinical experience underscores the large reduction in operative morbidity
that occurs from a mapping/sentinel node procedure in comparison to that
from a formal axillary dissection. The sentinel node technique is now being
used more widely and may be particularly useful in patients for whom the
indications for systemic adjuvant therapy are borderline, such as small
invasive tumors. It is also applicable to many patients with DCIS.
When a sentinel node or nodes are demonstrated to
be involved by tumor using hematoxylin and eosin (H & E) staining techniques,
a completion node dissection is indicated. The development of more sensitive
techniques to evaluate resected lymph nodes for micrometastasis, however,
has created new problems in staging management recommendations. Experience
at the Moffitt Cancer Center using cytokeratin immunohistochemical (CKIHC)
staining in addition to routine H & E stains in a group of 210 breast
cancer patients12 indicates that CKIHC staining of sentinel
nodes shifted the stage of 9.4% of them from stage I to stage II. The effect
was particularly strong in patients with tumor diameters <2.0 cm. Sixteen
of 30 such node-positive T0 and T1 patients (53.3%) were upstaged by CKIHC
staining.
The upstaging effect from more sensitive and sophisticated
diagnosis will increase apparent survival rates for both stages I and II
breast cancer by the "Will Rogers" effect. The clinical meaning of these
results, however, is unclear. This complicates decision making. Thus, we
now have to ask if a patient with DCIS only, ie, with no invasive component
in the primary cancer who has a "cytokeratin-only" nodal micrometastasis,
should receive adjuvant therapy.
Historical data from Pickren13 and Wilkerson
et al,14 who studied serial sections of uninvolved nodes using
H & E staining, suggest that the finding of microscopic-only nodal
disease did not affect survival outcomes. Prospective trials will be needed
to sort these issues out properly.
Patterns of Care
Two presentations at the 1998 ASCO meeting addressed
national patterns of care for patients with early breast cancer.
15,16
Joint studies of the American College of Surgeons and the American College
of Radiology included 17,931 patients with stage I and stage II breast
cancer treated at 827 institutions in 1994. One study evaluated management
of the axilla.
15 Ninety-three percent of the group underwent
axillary lymph node dissection (ALND), but this procedure was performed
less often in older patients, in those with T1a tumors, and in those with
favorable histologies (tubular, papillary, and mucinous carcinomas). Patients
not undergoing ALND were more likely to receive axillary radiation than
those who did (27% vs 12%, respectively,
P=0.0001), although only
30% of patients with >=10 nodal metastases received axillary radiation.
This proportion seems inappropriately low.
The second study evaluated treatment of the breast
itself. There remain significant regional variations in the methods used
to provide control of the primary tumor in the breast.16 Overall,
only 44% of studied patients underwent breast conservation therapy. This
option was followed mostly for patients with favorable prognostic indicators
for survival, despite guidelines and data from randomized trials indicating
that age, prognosis, and tumor type should not be used as selection criteria
for conservation therapy. This misunderstanding is a major factor responsible
for continued low national rates for breast-conserving therapy, and continued
education on these points is indicated.
Postmastectomy Radiation
Long a contentious issue, recommendations for postmastectomy
radiation in patients with fewer than four axillary nodes involved and
clear surgical tumor resection margins are in a state of flux. The issue
seemed to have been put to rest by the meta-analysis published by the Early
Breast Cancer Trialists’ Group in 1995.
17 This meta-analysis
concluded that "some of the local therapies for breast cancer had substantially
different effects on the rates of local recurrence, such as the reduced
recurrence with the addition of radiotherapy to surgery --
but there
were no definite differences in survival at 10 years."
The simultaneous publication of two well-conducted
prospective trials18,19 with long follow-up periods in premenopausal
patients with breast cancer who were at high risk of recurrence and who
received chemotherapy with CMF (cyclophosphamide, methotrexate, and fluorouracil)
± radiation therapy last year, however, indicated that not only
was local recurrence reduced but that survival rates were also improved
by radiation. This evidence has stimulated considerable debate. The M.D.
Anderson group reported at the 1998 ASCO meeting on 1,107 of their early
breast cancer patients who received FAC (fluorouracil, doxorubicin, and
cyclophosphamide) adjuvant therapy.20 Retrospective analysis
of outcomes between those who received radiation and those who did not
(ie, not an entirely randomized study) showed that pretreatment characteristics
of patients treated with or without radiation were similar. The risk of
local failure was higher in the FAC alone group (11%) vs that for FAC plus
radiation (5.6%), but there was no difference between the two groups in
disease-free or overall survival. At that institution, postmastectomy radiation
is not given to patients with fewer than 4 nodes. It is prescribed, of
course, for those at high risk of local recurrence, including stage III
and inflammatory cancer, positive resection margins, extensive extranodal
axillary disease, or >=10 involved axillary nodes. This presentation will
not put to rest the controversy regarding postmastectomy radiation in patients
who have 1 to 3 nodes involved. Many clinicians would support a study that
addresses this question directly. Meanwhile, discussions among medical
and radiation oncologists will continue.
Adjuvant Systemic Therapy
At long last, the final version of the first part of
the updated report of the Early Breast Cancer Trialists’ Collaborative
Group has been published,
4 although preliminary observations
concerning the 1995 Oxford Overview have previously been summarized.
21
Tamoxifen
The 1998 Lancet publication4 concerns
tamoxifen and presents an analysis of 37,000 women with breast cancer in
55 trials that randomized between treatment with tamoxifen or not. Nearly
8,000 of the women had ER-negative tumors, and the effects of tamoxifen
in these women are small. When the remaining women (18,000 with ER-positive
tumors and nearly 12,000 with untested tumors, of which approximately 8,000
would have been ER-positive) were evaluated, the proportional cancer recurrence
reductions produced during approximately 10 years of follow-up were 21%,
29%, and 47% for trials of one, two, and five years of adjuvant tamoxifen,
respectively. The corresponding mortality reductions were 12%, 17%, and
26%. The proportional mortality reductions were similar for women with
node-positive and node-negative disease, but the absolute mortality reductions
were greater in node-positive women. In the trials of approximately five
years of tamoxifen, the absolute improvements in 10-year survival were
10.9% for node-positive patients (61.4% vs 50.5%) and 5.6% for node-negative
patients (78.9% vs 73.3%). These benefits were largely independent of age,
menopausal status, and chemotherapy. The data now clearly show that tamoxifen
is beneficial in premenopausal women.
The proportional reductions in contralateral breast
cancer were 13% (SD 13), 26% (SD 9), and 47% (SD 9) in the trials of one,
two, or approximately five years of tamoxifen, respectively. The incidence
of endometrial cancer increased with longer duration of tamoxifen, but
the absolute decrease in contralateral breast cancer was approximately
twice as large as the absolute increase in endometrial cancer incidence.
No effect was reported in the incidence of colorectal, liver, or other
cancers, or of other major categories of cause of death.
Thus, in summary, beneficial effects of tamoxifen
are clearly evident in patients with ER-positive carcinomas that may be
node negative or node positive. Benefit is seen in premenopausal and postmenopausal
patients, and the effects are independent of chemotherapy. Greater effects
are seen with five years’ duration rather than with shorter durations of
tamoxifen adjuvant therapy. Studies of adjuvant therapy using toremifene,
as well as comparative studies of tamoxifen and anastrozole and the combination
of the two drugs, are underway. I assume that the capability of raloxifene
to control or delay systemic breast cancer growth and spread will also
be evaluated.
Cytotoxic Therapy
Perhaps the most commonly used cytotoxic adjuvant drug
programs for breast cancer used in practice are CMF, CAF, and CA. The original
or "classic" CMF program comprises 28-day cycles, with methotrexate and
fluorouracil given on days 1 and 8 and cyclophosphamide given orally on
days 1 through 14 of each cycle. At least in my geographic area, oncologists
often prescribe an admittedly better-tolerated all-intravenous (IV) schedule
given every three weeks. A "standard" course of adjuvant therapy consists
of 6 cycles of the "classic" CMF, but I often see recommendations for only
6 cycles of the q 3 week IV CMF approach.
The IV CMF program is less dose intensive than the
"classic" 28-day program, and studies in patients with advanced breast
cancer show better outcomes when the "classic" CMF schedule is used.22
Goldhirsch and colleagues from the International Breast Group23
have recently reviewed this information and point out that reduction of
doses has given inferior results in the adjuvant setting. The trials that
demonstrated a significant benefit from the addition of CMF to tamoxifen,
even in ER-positive tumors, used "classic" CMF. The authors suggest that
"adherence to the classical dose and schedule is recommended when CMF is
used in adjuvant therapy." I concur with this recommendation and suggest
that practitioners who now routinely use the IV CMF treatment program review
the data in the Goldhirsch article and consider changing their approach
to reuse the "classic" CMF schedule.
The preliminary results of the Intergroup trial of
CMF vs CAF with and without tamoxifen in high-risk, node-negative breast
cancer patients were featured in the 1998 ASCO plenary session.24
This study of 4,406 patients classified the patients into three risk categories:
high (tumor >=2 cm or ER- and PR-negative), low (tumor too small for biochemical
hormone receptor studies), and uncertain (tumors <2 cm and ER-negative
or PR-positive). Those in the uncertain group were assigned to the low-risk
group if the S-phase fraction was low and to the high-risk group if it
was high. The low-risk group was followed without adjuvant therapy. Table
4 summarizes the preliminary results.
Table 4. --
Differences in Five-Year Observed Survival (OS) and Disease-Free
Survival (DFS) in 2,691 High-Risk and 1,208 Low-Risk Patients with Node-Negative Breast
Cancer by Treatment |
| |
Low risk by: |
High risk: |
All Patients: |
Small tumor |
Low SPF |
CAF+ |
Low Risk |
High Risk |
CAF |
TAM |
CMF |
OS |
96% |
97% |
92% |
93% |
90% |
96% |
92% |
DFS |
89% |
88% |
85% |
87% |
82% |
89% |
85% |
| |
| SPF = S-phase fraction |
CAF was marginally superior to CMF in increasing
both disease-free and overall survival (one-sided P=0.03) but was
somewhat more toxic. Tamoxifen provided additional benefit only in ER-positive
patients. The subset of low-risk patients, chosen on the basis of tumor
size, receptor status, and S-phase fraction do well without adjuvant therapy,
thus validating the utility of these prognostic markers in node-negative
patients.
Preliminary results from another large intergroup
trial for breast cancer, this time in node-positive patients, were also
presented at the 1998 ASCO meeting.25 The study had two components.
The first was to evaluate whether dose-escalation of doxorubicin from 60
mg/m2 to 75 mg/m2 or 90 mg/m2 in combination
with 600 mg/m2 of cyclophosphamide for 4 cycles would improve
results. It did not. The second component of the trial evaluated the effects
of the addition of 4 cycles of paclitaxel given at 175 mg/m2
over 3 hours. Table 5 summarizes the results.
Table 5. --
Disease-Free Survival (DFS) and Observed Survival (OS) in 3,170
Breast Cancer Patients With Involved Lymph Nodes Treated by Cyclophosphamide Plus
Doxorubicin Alone (CA) or by CA Followed by Paclitaxel (CA-›T) |
| |
CA |
CA-›T |
P = |
DFS |
86% ± 1.2% |
90% ± 1.0% |
0.0077 |
OS |
95% ± 0.7% |
97% ± 0.6% |
0.0390 |
Evidence to date indicates that the sequential addition
of paclitaxel to CA adjuvant therapy for node-positive breast cancer is
well tolerated and significantly improves both disease-free and observed
survival. Underscoring the value of this advance is the fact that the "standard"
arm of the Intergroup adjuvant protocol #C9741 for patients with node-positive
disease is 4 cycles of CA followed by 4 cycles of paclitaxel given at three-week
intervals. The "test" treatment is the same drug regimen but given at two-week
intervals, thus testing the "dose-dense" approach.
HER-2/neu
An integrated symposium on HER-2/neu in breast cancer
was held at the 1998 annual ASCO meeting. Interest lies in the value of
measurements of overexpression of this gene in breast cancer tissue for
prognostic and predictive information, and also on the value of an anti-HER-2/neu
antibody as a treatment for breast cancer either alone or as an adjunct
to cytotoxic therapy.
HER-2/neu, also known as c-erb-B2, is found
to be overexpressed in approximately one third of breast cancer specimens
from women with localized breast cancer, and such expression is considered
to predict overall for a poor outcome.26 Presentation