Annuls of Oncology 3: 801-807, 1992.

© 1992 Kluwer Academic Publishers. Printed in the Netherlands.

Adljy vaunt therapy of primary breast camcer *

4th International Conference on Adjuvant Therapy of Primary Breast Cancer St. Gallen,
Switzerland

J. H. Glick,1 R. D. Gelber,2 A. Goldhirsch3 & H. J. Senn4

1
University of Pennsylvania Cancer Center, Philadelphia, PA; 2Dana Farber Cancer Institute, Boston, MA, U.S.A.; 3Ospedale Civico,
Lugano; "Kantonsspital Oncology Center, St. Gallen, Switzerland

Introduction statistically significant reductions in the annual rates of

both recurrence and of death produced by tamoxifen in
Adjuvant therapy for stage I and II breast cancer re- both node-negative and node-positive patients, by
mains in evolution as new data emerge from controlled ovarian ablation below age 50, and by combination
randomized trials and from the overview analysis by chemotherapy for patients below age 50 and for those
the Early Breast Cancer Trialists' Collaborative Group. between 50-69 years of age [4]. For tamoxifen and
The Fourth International Conference on Adjuvant combination chemotherapy, the avoidance of recur-
Therapy of Primary Breast Cancer, held in St. Gallen, rence is chiefly during the first 5 years, but the avoid-
Switzerland in February, 1992, brought together breast ance of mortality is highly significant during both the
cancer experts from all over the world to present their first and second 5 years, so the cumulative differences
most recent data and to discuss the remaining chal- in survival are large both at five and then at 10 years.
lenges that confront both clinicians and basic scientists. Indirect comparisons show that long-term tamoxifen
The data presented at the 1992 St. Gallen confer- (e.g., two years or even five years) is significantly more
ence allow us to revisit the treatment recommendations effective than shorter duration tamoxifen treatment
made at the 1985 and 1990 NIH Consensus Develop- programs. Between ages 50 and 69, direct comparisons
ment Conferences, as well as at the 1988 St. Gallen show that chemotherapy plus tamoxifen is superior to
meeting [1-3]. The timing of the 1992 St. Gallen Con- chemotherapy alone for both recurrence and mortality,
ference was fortunate in that the ten-year data from the and better than tamoxifen alone for recurrence. In
Early Breast Cancer Trialists' Collaborative Group women aged under 50, chemotherapy and ovarian
(hereafter referred to as the Overview) had recently ablation appear, by indirect comparison across two
been published [4]. Four distinct overviews were con- separate overviews, to be of comparable efficacy. The
ducted to evaluate whether adjuvant therapy with 30%-40% proportional risk reductions that can be
produced by combined chemo-endocrine therapy in
tamoxifen, chemotherapy, ovarian ablation or immuno-
middle age are similar for node-positive and for node-
therapy, either alone or as part of combined modality
negative patients, while the absolute improvement in
treatment, reduced the risk of recurrence and death.
ten-year survival is about twice as great for the node-
Eligible randomized trials included those started be-
positive patients [4]. Thus, the conclusions from the
fore January, 1985, in which the modality under study ten-year Overview analysis clearly influenced the treat-
was the only difference between treatments received by ment recommendations made at the 1992 St. Gallen
two randomized groups of patients. Data were available Conference, which will be discussed in a later section.
for 75,000 women randomized in 133 clinical trials. The timeliness of this Conference is illustrated by the
The specific positive findings that attracted attention remarkable lack of conflict from different investigators,
were based on 30,000 women in tamoxifen trials, 1800 who, using their own data and that from the ten-year
women below age 50 in ovarian ablation trials, and Overview, came to similar conclusions regarding the
11,000 women in combination chemotherapy trials. 1992 St. Gallen treatment recommendations.
It was clear that the ten-year Overview data present-
ed by Peto and colleagues, greatly influenced the par-
ticipants at the 1992 St. Gallen Conference. The Over-
view conclusions included demonstration of highly Conference highlights
* Published previously as Meeting Highlights: Adjuvant Therapy for
Primary Breast Cancer in Journal of the National Cancer Institute 84: The role of currently accepted and new prognostic fac-
1479-85, 1992. tors was a subject of active debate. There was broad
802

acceptance that rumor size, axillary lymph node status, able breast cancer.. The primary aim of this trial was to
hormone receptors and histopathology are important increase the frequency of breast conservation treatment
prognostic factors. Both Clark et al. [5], and Dressier et in patients with an initial tumor size >3 cm. In this
al. [6], stressed the importance of determining S-phase series of 227 patients, 82% had tumors measuring be-
by flow cytometry as an important prognostic factor for tween 3.0-5.0 cm at the time of entry on study, while
node-negative patients. While S-phase determination by only 18% of tumors initially measured greater than 5.0
flow cytometry is broadly accepted in the United States cm. Bonadonna administered 3-4 cycles of chemother-
as a routinely used prognostic variable for node-nega- apy prior to surgery. The surgical technique included
tive patients, Silvestrini from the NCI in Milan [7], did either lumpectomy or quadrantectomy for tumor sizes
not confirm the value of S-phase as determined by flow measuring less than 3 cm and mastectomy for tumors
cytometry for these patients. Instead, Silvestrini argued greater than 3 cm. Definitive radiotherapy was then ad-
that the 3H-thymidine labelling index was the most ministered. Although 91% of patients on this trial sub-
important indicator of relapse. Clark et al. [5] discussed sequently underwent breast conservation surgery, and
the prognostic significance of over-expression of the only 9% required mastectomy, Fisher observed that
tumor suppressor gene p53 in node-negative patients. 75%-80% of Bonadonna's patients would have been
The incidence of p53 mutations, as determined by eligible for initial lumpectomy in the NSABP B-06
immunohistochemistry, was observed in 50% of pa- trial. Thus, the use of neoadjuvant chemotherapy for
tients and independently predicted for early recurrence patients whose tumors measure 3.0 to 5.0 cm is not
in node-negative patients. Klijn et al. [8] discussed the routinely recommended, since these patients are gener-
prognostic value of the epidermal growth factor recep- ally candidates for lumpectomy without initial chemo-
tor (EGF-R), and concluded that EGF-R negative pa- therapy.
tients have improved disease-free and overall survival. Harris [13] discussed the integration of primary
However, they observed that the prognostic value and radiotherapy and adjuvant chemotherapy. In a small
best cut-off value of EGF-R remain open to question. retrospective series, they observed that significant
Despite the proliferation of new prognostic factors delays in the initiation of radiotherapy, in order to allow
each year, the St. Gallen Conference participants con- the delivery of chemotherapy before radiation, may in-
cluded that only the number of axillarly lymph nodes crease the rate of local recurrence in the treated breast.
involved, tumor size, age or menopausal status, histo- However, the small number of patients in this retro-
pathology, and hormone receptor status should be spective analysis make this conclusion tenuous at this
routinely used at this time. Baum [9] agreed with the time. Thus, the optimal sequencing of definitive radia-
routine use of these well-accepted prognostic factors. tion and chemotherapy remains an open question.
Data presented at the Conference continue to dem- Kurtz [14] also reviewed factors influencing the risk
onstrate that breast conservation therapy (lumpectomy of breast recurrence following lumpectomy and radia-
and radiation therapy) achieves equal ten-year overall tion. He noted that patients age less than 35 incurred a
survival as compared with modified radical mastec- significant increased risk of local recurrence as com-
tomy. Thus, lumpectomy and radiotherapy offer equal pared to older patients (p < 0.05). There was no in-
efficacy with less morbidity and are the preferred local fluence of menopausal status, location of the primary in
treatment for most patients with early stage breast can- the breast, tumor size, axillary node status or hormone
cer. Margolese [10], reporting the data from NSABP, receptors in predicting for an increased local breast
noted a 40% rate of local breast recurrence at nine recurrence. Patients with an extensive intraductal com-
years in patients who were treated with lumpectomy ponent did have a significantly increased rate of local
alone. However, there was no difference in overall sur- failure. Extensive intraductal component has been de-
vival between the lumpectomy alone as compared to fined by Harris [13] as infiltrating ductal carcinoma in
lumpectomy and radiotherapy patients. A very low which intraductal carcinoma is predominantly present
incidence of breast recurrence has been observed in in the tumor (>25%) and intraductal carcinoma is pre-
recent NSABP trials in which chemotherapy was used sent in sections of grossly normal adjacent breast tissue.
in addition to radiotherapy to the conserved breast Peto reviewed the data from the Early Breast Can-
(2.6% breast recurrences in the chemotherapy group cer Trialists' Collaborative Group [4[ and comment-
compared to 7.4% in the no treatment group) [11]. ed on the importance of these data as follows: the
For patients with tumor size <1.0 cm, the NSABP trial definitive effect of adjuvant therapy on ten year sur-
did not demonstrate a significant difference in local vival is clearly demonstrated; the extra mortality bene-
breast recurrence rate between lumpectomy alone as fit was seen between years five and nine with the addi-
compared to lumpectomy plus radiotherapy. Thus, tion of tamoxifen; ovarian ablation in patients age less
controversy remains as to whether there is a subgroup than 50 produced similar effects in the reduction of
of patients with smaller tumors (i.e., < 1 cm), who can recurrence and mortality to those seen with combina-
be treated with lumpectomy alone and spared the cost tion chemotherapy; tamoxifen was effective in women
and possible morbidity of radiation therapy. greater than age 50 in significantly improving both
Bonadonna [12] discussed the Milan trial using pri- relapse-free survival and overall survival for all women,
mary systemic (neoadjuvant) chemotherapy for oper- even those who are estrogen receptor poor; there was a
 803

definite reduction in mortality from hormonal treat- mained investigational and should not be used out-
ment in node-negative patients; the administration of side the context of the randomized controlled clinical
tamoxifen significantly reduced the incidence of con- trial.
tralateral breast cancers. The Overview data [4] indi- Tormey [18] reviewed the data from several random-
cate a highly significant trend towards a greater ized trials conducted by the Eastern Cooperative On-
therapeutic effect for long-term (greater than 2 years) cology Group. These trials particularly addressed the
tamoxifen. role of long-term tamoxifen for at least five years versus
Bonadonna then critically reviewed the endocrine shorter durations of tamoxifen (i.e., one year). The
data from the ten-year Overview [15]. He recommend- studies demonstrated that long-term tamoxifen (5
ed tamoxifen alone only for estrogen receptor positive years) was associated with a significantly longer time to
node-negative patients, and was not impressed by the recurrence than the shorter term tamoxifen (one year)
tamoxifen data for estrogen receptor poor patients. regimens (p < 0.02). However, there are no survival dif-
Bonadonna commented that, to improve treatment ferences noted as yet in these trials.
outcome, he recommended the use of both tamoxifen Fisher [11, 19] reviewed the NSABP data on node-
and chemotherapy as partners due to the known tumor positive patients who were defined as 'non-responsive
heterogeneity of breast cancer. For the majority of pa- to tamoxifen', based on definitions in older NSABP
tients requiring adjuvant therapy, either node-negative studies which included selection criteria using age and
or node-positive, Bonadonna recommended sequential progesterone receptor status. Doxorubicin and cyclo-
treatment with chemotherapy for six months followed phosphamide administered every 21 days for four
by tamoxifen for at least five years. Baum [9] remarked cycles achieved equivalent disease-free and overall sur-
that the ten-year Overview data provided clear evi- vival to conventional CMF adjuvant chemotherapy. In
dence that chemotherapy was standard treatment for 'tamoxifen-responsive', estrogen receptor positive pa-
premenopausal patients with positive axillary lymph tients doxorubicin and cyclophosphamide were admin-
nodes, while tamoxifen remained the standard therapy istered concurrently with tamoxifen (ACT) and were
for postmenopausal patients with positive lymph compared to tamoxifen alone. The ACT regimen
nodes. However, Baum admitted that the Overview achieved significantly improved five-year disease-free
data now raised uncertainties in his mind about the role survival (70%) as compared to tamoxifen alone (60%)
of adjuvant chemotherapy in node-positive postmeno- (p = 0.002), as well as significantly improved five-year
pausal patients. overall survival (ACT 87% vs. 76% with tamoxifen
Gelber [16] recommended that the arithmetic con- alone, p = 0.002). The data from this NSABP trial,
struct of the Overview be broken up to distinguish when viewed in the context of the ten-year Overview
single modality from combined modality treatment analysis, was influential in the proactive treatment re-
effects. By using trials which evaluated treatments commendations for chemotherapy in postmenopausal
actually given (for example, chemoendocrine therapy patients made at the St. Gallen Conference. Although
versus chemotherapy or endocrine therapy alone), he the Conference participants recognized that the major-
concluded that chemoendocrine therapy appears to ity of individual trials have not shown a survival benefit
provide the best chance of success in all patients (both for adjuvant chemotherapy in the postmenopausal,
pre- and post-menopausal); although even with all of node-positive group, the Overview analysis was impor-
the Overview results, additional trials are needed to tant in the decision to recommend chemotherapy in
define the patients who benefit and the best way to this subgroup.
combine modalities. He stressed the importance of the The NSABP trials in node-negative patients were
Overview for stimulating international collaboration also reviewed by Fisher [11, 19] who presented the
and identifying questions for future clinical trials which seven-year data from their B-13, which randomized pa-
will improve patient care. tients with ER-negative tumors to methotrexate and
The role of high-dose chemotherapy in high-risk 5-fluorouracil versus a no treatment control arm. At
node-positive patients with >ten positive axillary nodes seven years, there remains a highly significant improved
was discussed by Abeloff [17]. Using a dose-intensive disease-free survival on the chemotherapy arm for both
16-week regimen developed at Johns Hopkins, Abeloff pre- and postmenopausal patients. Although no differ-
reported a 53% disease-free survival with a median ence in overall survival for all patients entered on this
follow-up of 40 months in this subgroup of patients. trial was observed, in the subgroup of patients >50
The data from Peters at Duke was also reviewed by years old, there is a significant improvement in overall
Abeloff. With a median follow-up of two years in 85 survival (p = 0.004). An additional important observa-
eligible patients treated with standard adjuvant chemo- tion was made by Fisher concerning tumors measuring
therapy, followed by high-dose chemotherapy and bone <1 cm in size. With methotrexate and 5-fluorouracil,
marrow transplantation, the Duke group observed 72% 90% of these patients were disease-free at seven years
five-year event-free survival. Although these data are compared to only 75% on the control arm (p < 0.05).
provocative, there was general agreement at the St. The relatively poor prognosis for patients with smaller
Gallen meeting that the role of adjuvant high-dose tumors was at variance with the data presented at 1990
chemotherapy and bone marrow transplantation re- NIH Consensus Conference [2|. Patient selection fac-
804

tors for those entering this NSABP trial may explain 0.9% of chemotherapy-treated patients. In addition, the
this apparent discrepancy. incidence of doxorubicin-related cardiotoxicity was
Fisher also presented updated data on the tamoxifen only 0.8%, while radiotherapy-related lung fibrosis was
versus placebo trial in node-negative, estrogen receptor noted in 3.4% of patients (primarily diagnosed by chest
positive patients [11, 19]. Significant improvement in x-ray).
both disease-free and overall survival at seven years Gelber et al. [23], discussed how to compare the
was noted for patients on the tamoxifen arm. The bene- quality of life of breast cancer patients in clinical trials.
fit for tamoxifen was also seen for patients with smaller They recommended that this is best done within ran-
tumors <1.0 cm. In this latter subgroup of patients, domized trials using the Q-TWiST technique based on
only 78% of the placebo-treated patients were disease- time with and without specific types of toxic effects of
free at seven years as compared to 82% of the tamo- therapy or with and without symptoms of recurrent dis-
xifen group. However, the number of patients in each ease as end points. They discussed methods to define
arm was relatively small (approximately 200 patients), the emotional adjustment which concerns almost every
since quantitative estrogen and progesterone receptors patient who has to be faced with the diagnosis of breast
were required for eligibility. Thus, Fisher was unable to cancer. Their methodology facilities the comparison of
confirm Rosen's data on patients with tumors measur- treatment benefits and subjective side-effects from the
ing < 1.0 cm, who were reported to have a 86% twenty- patient's point of view. This methodology could be
year disease-free survival [20]. It should also be noted adopted in current and future trials of adjuvant therapy,
that the NSABP did not include patients with < 1.0 cm particularly where the differences in treatment out-
tumors which were detected only by mammography. come are expected to be of small biological value.
Jordan et al. [21] used a laboratory model to de- Hillner [24] discussed the financial costs, benefits
scribe the survival advantage observed by patients tak- and the role of patient risk preferences. He described a
ing adjuvant tamoxifen. In their model, tamoxifen model that used the results of available randomized
initially inhibits estrogen-stimulated growth of MCF-7 controlled trials in node-negative patients to assess the
breast cancer cells transplanted into athymic mice. potential clinical and financial effects of the universal
However, treatment of animals transplanted with use of adjuvant chemotherapy. Using baseline assump-
MCF-7 tumors with tamoxifen for up to one year tions, Hillner found that chemotherapy increases qual-
results in tamoxifen-stimulated tumor growth. At four ity adjusted life expectancy and survival at a cost com-
years, these tamoxifen-stimulated tumors still main- parable to most medical interventions. The model high-
tained estrogen receptors, but the replacement of tamo- lighted the need for greater study of patient preferences
xifen with physiological circulating levels of estradiol and the need for accurate cost-accounting for onco-
causes the immediate and total regression of estab- logic therapies and the role of patient preferences.
lished tumors. Jordan suggested that during adjuvant
treatment with tamoxifen, tumor cells are initially pre-
vented from growing during the first year but then Conclusions: The 1992 St. Gallen treatment recom-
clones of tamoxifen-stimulated tumor cells eventually mendations
emerge. The benefit from tamoxifen comes when treat-
ment is stopped, and the circulating estradiol kills The 1985 and 1990 NIH Consensus Conferences, as
tamoxifen-dependent disease. This provocative labora- well as the recent 10-year Overview from the Early
tory observation has not yet been tested in clinical Breast Cancer Trialists' Collaborative Group, present a
trials, but remains a new avenue of investigation. How- wealth of data indicating that adjuvant chemotherapy
ever, it should be noted that when tamoxifen treatment and hormonal therapy are effective treatments for
is stopped in the metastatic disease setting, tumor breast cancer patients with both positive and negative
growth occurs, and this may be due to the large tumor axillary lymph nodes [1, 2, 4]. However, there remains
volume seen in this situation. considerable debate regarding the use of adjuvant ther-
The long-term toxicities of adjuvant chemotherapy apy in many common situations in clinical practice. Al-
(CMF plus or minus doxorubicin) in 677 women with though optimal therapy has not been defined for any
stage II breast cancer were reviewed by Valagussa et al. subset of patients, it is unfortunate that less than 5% of
[22]. Doxorubicin was given to 455 patients, while all such patients in the United States actually are
breast radiotherapy after conservative surgery was entered into a clinical trial of adjuvant therapy. This is
given concurrently with chemotherapy in 291 patients. also true for most of Europe. Thus, practicing oncolo-
With a median follow-up of 52 months, virtually no gists and their patients are continually faced with a
long-term leukopenia was noted, while 2% of patients therapeutic dilemma as to which node-negative pa-
had a hemoglobin less than 12 grams. Venous throm- tients should be observed and which should be treated
bosis was observed in only 0.6% of patients undergoing with adjuvant therapy outside the setting of a clinical
chemotherapy. However, drug-related amenorrhea was trial. In addition, much controversy remains regarding
a common finding and was observed in 69% of pa- which subsets of node-positive patients should receive
tients. There was no significant increase in the inci- chemotherapy, hormonal manipulation, or both modal-
dence of second neoplasms, which were observed in ities. Outside the context of a clinical trial, and based
 805

on the research data presented at the 1992 St. Gallen logic and nuclear grade), the option of chemotherapy
Conference, which included the recently available data could be considered. The decision to use adjuvant ther-
from the 10-year Overview, we can summarize the apy in a patient with small tumors should follow a
treatment recommendations made at the 1992 St. thorough discussion with the patient regarding the like-
Gallen meeting (Tables 1-3). ly risk of recurrence without adjuvant therapy, the
expected reduction in risk with adjuvant therapy, toxic-
ities of therapy, and the impact on quality of life. Even
Node negative small absolute benefits in terms of reduction of disease
recurrence and mortality may be viewed by patients as
Node-negative: Minimal/low risk an important benefit. In that case, adjuvant therapy,
particularly with tamoxifen, for patients with these
Within the node-negative group of patients, a minimal/ small tumors would be indicated.
low-risk category can be defined, which includes pa-
tients with non-invasive tumors (ductal carcinoma in Node-negative: Good risk
situ), incidentally discovered small (<1 cm) invasive
tumors detected by mammography or by microscopic A good risk subgroup was defined as those patients
examination of tissue removed because of benign with approximately an 85%-90% chance of remaining
breast disease or because of in situ breast carcinoma. In disease-free at five years. This category includes pa-
addition, patients with tubular, colloid (mucinous), and tients who meet all the following criteria: tumor size > 1
papillary tumor types are included in this subgroup, cm but < 2 cm; and whose tumors are estrogen recep-
because of their low-risk of recurrence, particularly if tor-positive. Patients in this category would also have
their tumor size is less than 2 cm. For patients de- low histologic grade and low nuclear grade tumors.
scribed in this minimal/low risk category, observation Based on the Overview data there was remarkable con-
without adjuvant therapy remains the current treatment sensus that this group of good risk patients should be
recommendation (Table 1). However, the benefits of treated with tamoxifen (Table 1). This recommendation
tamoxifen for this subgroup of patients remain to be applied to both premenopausal as well as postmeno-
determined by ongoing clinical trials. For example, at pausal patients, and included the elderly patient in
this time, tamoxifen is not recommended for women this proactive treatment recommendation. Tamoxifen
with ductal carcinoma in situ outside the context of a should serve as the control arm for future clinical trials
clinical trial. If, after explaining the potential risks and in the good risk subgroup. Thus, for the first time, a
benefits from the administration of tamoxifen to a given general recommendation to use adjuvant tamoxifen
patient in this subgroup, the individual patient decides was made for the good risk subgroup of node-negative
that tamoxifen provides perceived benefit, then the patients.
administration of tamoxifen in this setting is not unrea-
sonable, and remains a viable option. Table 2. 1992 St. Gallen treatment recommendations.

Table I. 1992 St. Gallen treatment recommendations. Node-negative breast cancer: high-risk
Premenopausal Treatment
Node-negative breast cancer Receptor positive chemotherapy ± tamoxifen
Receptor negative chemotherapy
Minimal/low dose Good risk
Postmenopausal Treatment
Premenopausal Nil vs. tamoxifen Tamoxifen Receptor positive Tamoxifen ± chemotherapy
Postmenopausal Receptor negative Chemotherapy ± tamoxifen
Nil vs. tamoxifen Tamoxifen
Elderly (> 70 years) Tamoxifen Elderly (> 70 years) Tamoxifen (chemotherapy if tolerated)

The 1990 NIH Consensus Panel concluded that, in Node-negative: High risk
general, patients with rumors < 1 cm in diameter have
an excellent prognosis with less than a 10% recurrence A high risk subgroup of node-negative patients also can
rate at ten years, and adjuvant therapy was, therefore, be defined on the basis of well-known and accepted
not indicated for this subgroup of patients [2]. The data prognostic factors. This group includes patients with
presented at the 1992 St. Gallen meeting was less de- unequivocally estrogen receptor negative tumors which
finitive and more controversial regarding the benefits of are > 1 cm in size, estrogen receptor positive tumors > 2
adjuvant therapy for patients with tumors measuring cm in diameter, and nuclear grade III (poor) tumors.
< 1 cm in size. Consequently, the 1992 St. Gallen treat- The presence of any one of these unfavorable prognos-
ment recommendations leave open the options of ob- tic parameters justifies the use of adjuvant therapy out-
servation versus five years of tamoxifen as potential side the setting of a clinical trial. Additional prognostic
treatments for patients with tumors of this size. It factors such as high S-phase as determined by flow
should also be noted that for an individual patient (e.g., cytometry can be added to the prognostic profile to
0.9 cm tumor, estrogen receptor negative, high histo- guide the clinician and the individual patient. In addi-
806

tion to the consideration of these well-known prognos- gen receptor negative, adjuvant chemotherapy is
tic factors, the physician must evaluate the patient's indicated as standard treatment. The role of
general medical condition and psychosocial status be- tamoxifen when added to chemotherapy in these
fore deciding on a therapeutic recommendation. receptor negative patients remains controversial,
Therefore, virtually all patients in the high-risk and is under current investigation.
group should receive some form of adjuvant systemic 3. For the elderly patient (greater than 70 years of
therapy (Table 2). All of the recent major clinical trials age) in the high risk, node-negative group, tamo-
demonstrate significant prolongation of disease-free xifen appears to be of benefit irrespective of
survival with either tamoxifen or chemotherapy for this estrogen receptor status. However, it can be en-
subgroup of patients. Evidence of significant prolonga- visioned that chemotherapy could be adminis-
tion of overall survival is beginning to emerge from tered to selected older receptor-negative patients
these individual trials as well. Moreover, the ten-year with excellent performance status and in good
Overview data, while not subdivided on the basis of general medical condition, who are at high risk
risk factors, provides strong evidence that node-nega- for recurrence despite their node-negative dis-
tive patients should receive some form of adjuvant ' ease. In this subset of patients, full-dose chemo-
therapy. therapy might well be indicated.
It is now apparent that all patients with node-nega-
Table 3. 1992 St. Gallen treatment recommendations. tive breast cancer should have an oncologic consulta-
tion at the time of diagnosis, so that the patient and her
Node-positive breast cancer physician can discuss the patient's prognostic profile
and the available therapeutic options, as well as the risk
Hormone receptor Hormone receptor versus benefit ratio. Although the patient should be
negative positive
urged to participate in one of the many available clini-
Premenopausal Chemotherapy Chemotherapy cal trials, it must be recognized that the vast majority of
± tamoxifen patients are treated outside the context of a clinical
Postmenopausal Chemotherapy Tamoxifen ± trial. In this situation, a physician should recommend
± tamoxifen chemotherapy adjuvant chemotherapy ± tamoxifen for all high-risk
Elderly (> 70 years) ? chemotherapy tamoxifen
node-negative patients, tamoxifen for the overwhelm-
ing majority of patients in the good risk category, and
observation versus tamoxifen for patients in the mini-
The 1992 St. Gallen treatment recommendations for mal/low-risk subgroup. Thus, the treatment recom-
node-negative, high risk breast cancer patients include: mendations made at St. Gallen in 1992 support the
1. All premenopausal, node-negative, high-risk pa- general recommendations made in the controversial
tients should receive adjuvant chemotherapy ir- 1988 NCI Clinical Alert [25].
respective of their estrogen-receptor status. For
receptor positive patients in this subgroup, the
combination of chemotherapy plus tamoxifen
also must be considered, but needs to be com- Node positive
pared prospectively to chemotherapy alone for
this subset of patients. Outside the context of a All patients with node-positive breast cancer should re-
clinical trial, for the premenopausal receptor ceive some form of adjuvant therapy, according to the
positive, node-negative patient, either chemother- treatment recommendations made at the 1992 St.
apy alone, chemotherapy plus concurrent tamo- Gallen Conference (Table 3).
xifen, or chemotherapy followed sequentially by 1. For premenopausal, node-positive patients who
tamoxifen are appropriate treatment options. For are either hormone receptor positive or negative,
the receptor-negative, premenopausal patient, ad- adjuvant chemotherapy remains the standard of
juvant chemotherapy remains the treatment of care and should be administered to all patients. In
choice at this time. the hormone-receptor-negative patient in this
2. For postmenopausal, node-negative, high-risk pa- category, chemotherapy alone is the treatment of
tients, the primary therapeutic modality depends choice. For the hormone-receptor-positive, pre-
on estrogen receptor status. For patients in this menopausal patient, the addition of tamoxifen to
category who are estrogen receptor positive, chemotherapy remains a viable treatment option.
tamoxifen remains the mainstay of treatment. In these patients, it is not known whether tamo-
There is increasing evidence both from individual xifen should be administered concurrently with
trials and from the Overview data, that tamoxifen chemotherapy or sequentially after the chemo-
combined with chemotherapy for this subgroup therapy is completed. Although ovarian ablation
of patients may have additional benefit. Thus, the achieves a relative reduction in the annual odds of
combination of tamoxifen plus chemotherapy is a recurrence and mortality of approximately 25%,
reasonable alternative to tamoxifen alone for this the 1992 St. Gallen Conference did not recom-
subgroup. For postmenopausal patients, in the mend the routine use of ovarian ablation outside
high-risk, node-negative category, who are estro- the context of a clinical trial. The role of ovarian
 807

ablation, particularly achieved using LH-RH ago- 3. Glick JH. Adjuvant therapy for breast cancer. JNCI 1988; 80:
nists such as Zoladex, either by itself or compared 471-5.
to chemotherapy alone, or in addition to chemo- 4. Early Breast Cancer Trialists' Collaborative Group. Systemic
treatment of early breast cancer by hormonal, cytotoxic, or
therapy for premenopausal, receptor-positive, immune therapy: 133 randomized trials involving 31,000 re-
node-positive patients is under active investigation. currences and 24,000 deaths among 75,000 women. Lancet
2. For postmenopausal, node-positive, hormone-re- 1992; 339: 1-5 and 71-85.
ceptor-positive patients, tamoxifen remains the 5. Clark GM, McGuire WL. Prognosis of breast cancer patients:
accepted treatment. However, data from both the How to use what? Recent Results in Cancer Research 1992; in
press.
NSABP and the ten year Overview provide in- 6. Dressier LG. DNA flow cytometry measurements: Its clinical
creasing evidence of the added benefit of chemo- impact in primary breast cancer. Recent Results in Cancer
therapy when given in addition to tamoxifen for Research 1992; in press.
this subgroup of patients. Again, the proper 7. Silverstrini R. Review of proliferative variables and their
sequencing of chemotherapy and tamoxifen re- predictive value. Recent Results in Cancer Research 1992; in
press.
mains to be defined.
8. Klijn JGM, Berns PM, Bontenbal M et al. Critical review of
3. For postmenopausal, node-positive, hormone-re- growth factors as clinical tools. Recent Results in Cancer
ceptor-negative patients, a clear consensus favor- Research 1992; in press.
ing adjuvant chemotherapy emerged. The strong 9. Baum M. What did we learn from the international overview
recommendation for adjuvant chemotherapy results about the effects of chemotherapy? Recent Results in
Cancer Research 1992; in press.
made at the 1992 St. Gallen Conference repre- 10. Margolese G. Selection and technique for lumpectomy. Recent
sents for the first time a broad international con- Results in Cancer Research 1992; in press.
sensus on the use of chemotherapy as standard 11. Fisher B, Wickerham DL, Redmond C. Recent developments
treatment for this group of patients. In this subset in the use of systemic adjuvant therapy for the treatment of
of receptor-negative patients, there is increasing breast cancer. Sem in Oncol 1992; 19: 263-77.
evidence that tamoxifen may be added to chemo- 12. Bonadonna G. Primary chemotherapy for resectable breast
cancer. Recent Results in Cancer Research 1992; in press.
therapy to improve disease-free and overall sur- 13. Harris J. How to combine adjuvant chemotherapy with radia-
vival. Therefore, the addition of tamoxifen to tion therapy. Recent Results in Cancer Research 1992; in press.
chemotherapy for this subgroup of patients re- 14. Kurtz JM. Factors which predict breast relapse. Recent Results
mains an increasingly viable treatment option to in Cancer Research 1992; in press.
be studied in clinical trials. 15. Bonadonna G. What did we learn from the results of the inter-
national overview about the effects of endocrine therapy?
4. For the elderly (greater than 70 years old), node- Recent Results in Cancer Research 1992; in press.
positive, hormone-receptor-positive patient, tamo- 16. Gelber RD, Goldhirsch A. From the overview to the patient:
xifen clearly remains the standard of care. For the How to interpret meta-analysis data. Recent Results in Cancer
elderly patient with node-positive, hormone-re- Research 1992; in press.
17. Abeloff MD. High dose chemotherapy for high risk breast can-
ceptor-negative breast cancer, the option of adju- cer? Recent Results in Cancer Research 1992; in press.
vant chemotherapy must be considered if the 18. Tormey DC. New aspects of chemoendocrine therapy. Recent
patient's general medical condition warrants the Results in Cancer Research 1992; in press.
administration of full-dose chemotherapy. Alter- 19. Fisher B. The recent NSABP experience and strategies. Recent
natively, tamoxifen could be administered to this Results in Cancer Research 1992; in press.
subset of receptor-negative elderly patients. 20. Rosen PP, Groshen S, Saigo PE et al. Pathological prognostic
factors in stage 1 (T2 No Mo) and stage II (T, N, Mo) breast
It now appears in 1992 that almost all patients with carcinoma. J Clin Oncol 1989; 7: 1239-51.
clinical and pathologic stage I and II breast cancer, 21. Jordan VC, Wolf DM. Laboratory model to describe the sur-
regardless of age, menopausal, nodal, or receptor sta- vival advantage observed by patients taking adjuvant tamoxifen
tus, will benefit from some form of adjuvant chemo- therapy. Recent Results in Cancer Research 1992; in press.
22. Valagussa P, Zambetti M, Bonadonna G. Long-term toxicities
therapy and/or hormonal therapy both in terms of im- of adjuvant chemotherapy. Recent Results in Cancer Research
proved disease-free survival, as well as overall survival. 1992; in press.
Thus, the 1992 St. Gallen treatment recommendations 23. Gelber RD, Cole BF, Goldhirsch A for the International Breast
confirm and expand on the guidelines suggested at the Cancer Study Group. How to compare quality of life of breast
time of the last St. Gallen Conference in 1988 [3|. Al- cancer patients in clinical trials. Recent Results in Cancer
though further clinical investigation is imperative, it is Research 1992; in press.
24. Hillner BE. Financial Cost, benefits and the role of patients
now incumbent upon each physician to apply the preferences. Recent Results in Cancer Research 1992; in press.
known adjuvant therapeutic regimens broadly, so as 25. Clinical Alert, National Cancer Institute, Bethesda, Maryland,
not to deny any patient the specific benefits of adjuvant May, 1988
therapy that already have been demonstrated.
Received 31 July 1992; accepted 5 August 1992.

References Correspondence to:

John H. Glick, M.D.
1. NIH Consensus Conference. Adjuvant chemotherapy for University of Pennsylvania Cancer Center
breast cancer. JAMA 1985; 254: 3461-3. 6 Penn Tower
2. NIH Consensus Conference. Treatment of early-stage breast 3400 Spruce Street
cancer. JAMA 1991; 265: 391-5. Philadelphia, PA 19104, U.S.A.