Prognostic and predictive factors in early-stage breast cancer.

Abstract

Breast cancer is the most common malignancy among American women. Due to increased screening, the majority of patients present with early-stage breast cancer. The Oxford Overview Analysis demonstrates that adjuvant hormonal therapy and polychemotherapy reduce the risk of recurrence and death from breast cancer. Adjuvant systemic therapy, however, has associated risks and it would be useful to be able to optimally select patients most likely to benefit. The purpose of adjuvant systemic therapy is to eradicate distant micrometastatic deposits. It is essential therefore to be able to estimate an individual patient's risk of harboring clinically silent micrometastatic disease using established prognostic factors. It is also beneficial to be able to select the optimal adjuvant therapy for an individual patient based on established predictive factors. It is standard practice to administer systemic therapy to all patients with lymph node-positive disease. However, there are clearly differences among node-positive women that may warrant a more aggressive therapeutic approach. Furthermore, there are many node-negative women who would also benefit from adjuvant systemic therapy. Prognostic factors therefore must be differentiated from predictive factors. A prognostic factor is any measurement available at the time of surgery that correlates with disease-free or overall survival in the absence of systemic adjuvant therapy and, as a result, is able to correlate with the natural history of the disease. In contrast, a predictive factor is any measurement associated with response to a given therapy. Some factors, such as hormone receptors and HER2/neu overexpression, are both prognostic and predictive.

Full text

Page 1

Prognostic and Predictive Factors in Early-Stage
Breast Cancer
MARY CIANFROCCA, LORI J. GOLDSTEIN
Fox Chase Cancer Center, Philadelphia, Pennsylvania, USA
Key Words. Breast cancer · Prognostic factors · Predictive factors · Adjuvant therapy
ABSTRACT
Breast cancer is the most common malignancy
among American women. Due to increased screening, the
majority of patients present with early-stage breast can-
cer. The Oxford Overview Analysis demonstrates that
adjuvant hormonal therapy and polychemotherapy
reduce the risk of recurrence and death from breast can-
cer. Adjuvant systemic therapy, however, has associated
risks and it would be useful to be able to optimally select
patients most likely to benefit. The purpose of adjuvant
systemic therapy is to eradicate distant micrometastatic
deposits. It is essential therefore to be able to estimate an
individual patient’s risk of harboring clinically silent
micrometastatic disease using established prognostic fac-
tors. It is also beneficial to be able to select the optimal
adjuvant therapy for an individual patient based on
established predictive factors. It is standard practice to
administer systemic therapy to all patients with lymph
node-positive disease. However, there are clearly differ-
ences among node-positive women that may warrant a
more aggressive therapeutic approach. Furthermore,
there are many node-negative women who would also
benefit from adjuvant systemic therapy. Prognostic fac-
tors therefore must be differentiated from predictive fac-
tors. A prognostic factor is any measurement available at
the time of surgery that correlates with disease-free or
overall survival in the absence of systemic adjuvant ther-
apy and, as a result, is able to correlate with the natural
history of the disease. In contrast, a predictive factor is
any measurement associated with response to a given
therapy. Some factors, such as hormone receptors and
HER2/neu overexpression, are both prognostic and
predictive. The Oncologist 2004;9:606-616
The Oncologist 2004;9:606-616
www.TheOncologist.com
Correspondence: Mary Cianfrocca, D.O., Fox Chase Cancer Center, 333 Cottman Ave, Philadelphia, Pennsylvania 19111, USA.
Telephone: 215-728-2974; Fax: 215-728-3639; e-mail: M_Cianfrocca@fccc.edu
publication May 21, 2004. ©AlphaMed Press 1083-7159/2004/$12.00/0
Received March 11, 2004; accepted for
INTRODUCTION
Breast cancer is the most common malignancy among
American women, with more than 200,000 new cases
diagnosed each year [1]. In recent years mortality from
breast cancer has declined in the U.S., likely as a result of
more widespread screening resulting in earlier detection as
The
Oncologist
Breast Cancer
®
LEARNING OBJECTIVES
After completing this course, the reader will be able to:
1. Differentiate between prognostic and predictive factors in early-stage breast cancer.
2. Identify prognostic factors used to determine the risk of recurrence and death for a patient with early-stage breast cancer.
3. Identify predictive factors used to determine the optimal therapy for a patient with early-stage breast cancer.
Access and take the CME test online and receive 1 hour of AMA PRA category 1 credit at CME.TheOncologist.com
CME
CME
by on October 17, 2009
www.TheOncologist.com
Downloaded from

Page 2

Cianfrocca, Goldstein
607
well as advances in the adjuvant treatment of early-stage
disease [2]. As a result of increased screening, the majority
of patients now present with early-stage breast cancer. In
1995, 56.2% of all breast cancer cases were stage 0 or 1
compared with 42.5% in 1985 [3]. The Oxford Overview
Analysis demonstrates that adjuvant hormonal therapy and
polychemotherapy reduce the risk of both recurrence and
death from breast cancer [4, 5]. Adjuvant systemic therapy,
however, does have associated risks, and it would therefore
be useful to be able to optimally select patients who are
most likely to benefit. Prognostic factors may select
patients most likely to recur without adjuvant therapy and
therefore potentially benefit from therapy. In addition, pre-
dictive factors may identify the appropriate therapy for an
individual patient.
ASSESSMENT OF RECURRENCE RISK
In the late 1800s, Halsted popularized the radical mastec-
tomy based on his belief that breast cancer spreads in an orga-
nized fashion, initially via the skin and regional lymphatics
and then, at a later stage, hematogenously to other organs.
Unfortunately, however, only 12% of patients treated with a
radical mastectomy survived 10 years. The poor outcome
with the Halstedian approach, as well as the observation that
20%-30% of node-negative patients ultimately develop
metastatic disease, led to the currently held micrometastatic
paradigm. This paradigm asserts that many patients with
early-stage disease have distant micrometastatic disease pre-
sent at the time of diagnosis, putting them at risk for the later
development of overt metastatic disease [6].
The purpose of adjuvant systemic therapy is to eradicate
these distant micrometastatic deposits. It is essential there-
fore to be able to estimate an individual patient’s risk of har-
boring clinically silent micrometastatic disease using
established prognostic factors. It is also beneficial to be able
to select the optimal adjuvant therapy for an individual
patient based on established predictive factors. Prognostic
factors therefore must be differentiated from predictive fac-
tors. A prognostic factor is any measurement available at the
time of surgery that correlates with disease-free or overall
survival in the absence of systemic adjuvant therapy and, as
a result, is able to correlate with the natural history of the dis-
ease. In contrast, a predictive factor is any measurement
associated with response to a given therapy. Some factors,
such as hormone receptors and HER2/neu overexpression,
are both prognostic and predictive.
It is currently standard practice to administer systemic
therapy to all patients with lymph node-positive disease.
However, there are clearly differences among node-positive
women that may warrant a more aggressive therapeutic
approach. Furthermore, there are many node-negative women
who would benefit from adjuvant systemic therapy. Clark
addressed the issue of prognostic and predictive factors and
suggested three main reasons to justify their use [7]. The
first reason is to identify patients with good prognoses for
whom adjuvant systemic therapy would not provide a large
enough benefit to warrant the risks. The second is to iden-
tify patients whose prognosis is poor enough to justify a
more aggressive adjuvant approach, and the third is to
select patients whose tumors are more or less likely to ben-
efit from different forms of therapy. Prognostic factors that
are considered to be independent variables include lymph
node status, tumor size, and estrogen/progesterone receptor
(ER/PR) status. Additional factors include grade, presence
of lymphovascular invasion, age, and ethnicity. Certain bio-
logic factors, including ER/PR and HER2/neu, are both
prognostic and predictive.
PROGNOSTIC FACTORS
Axillary Nodal Status
The most significant prognostic indicator for patients
with early-stage breast cancer is the presence or absence of
axillary lymph node involvement. Furthermore, there is a
direct relationship between the number of involved axillary
nodes and the risk for distant recurrence [8, 9]. For sim-
plicity, however, most clinical trials stratify patients based
on four nodal groups that are based on National Surgical
Adjuvant Breast and Bowel Project (NSABP) data: nega-
tive nodes, 1-3 positive nodes, 4-9 positive nodes, and 10 or
more positive nodes. The 5-year survival for patients with
node-negative disease is 82.8% compared with 73% for 1-
3 positive nodes, 45.7% for 4-12 positive nodes, and 28.4%
for ≥13 positive nodes [10]. These data demonstrate that the
risk of recurrence is significant enough with lymph node-
positive disease to warrant adjuvant systemic therapy since,
generally, a future risk of distant recurrence of 20% or
greater is regarded significant enough to consider the risks
of therapy. For lower-risk patients, especially those who are
node negative, an individualized assessment utilizing other
prognostic factors must be performed.
Traditionally, the status of the axilla has been assessed
by a standard axillary dissection in which level I and level
II lymph nodes were removed. Recently, the use of sentinel
node (SN) biopsy has become more common. SN biopsy
was first used to stage malignant melanoma [11]. The ini-
tial study of this technique in breast cancer was reported by
Giuliano et al. using the blue dye method [12]. SN were
identified in 65% of patients and accurately staged the
axilla in 96% of those patients. More recent studies using a
combination of blue dye and radiolabeled colloid have
achieved detection rates of greater than 95% [13].
by on October 17, 2009
www.TheOncologist.com
Downloaded from

Page 3

608
Prognostic and Predictive Factors in Early-Stage Breast Cancer
Although the ability of an experienced surgeon to accu-
rately stage the axilla with SN biopsy is accepted, multiple
questions remain, including the most suitable method to iden-
tify the SN as well as the optimal pathologic method to assess
the SN for involvement. Serial sectioning of each SN
increases the sensitivity as does the use of immunohisto-
chemistry (IHC) for histologically negative lymph nodes. The
significance, however, of occult micrometastases found by
IHC alone remains controversial [14-16]. A recent retrospec-
tive review with long-term follow-up demonstrated an
increased risk of recurrence and breast cancer-related death in
women who had occult or micrometastatic tumor deposits in
their axillary lymph nodes [17]. Similar results were observed
in the International Breast Cancer Study Group Trial V of
1,275 node-negative women randomly assigned to a single
cycle of perioperative cyclophosphamide, methotrexate, and
5-flourouracil (CMF) versus no chemotherapy [18]. The axil-
lary nodes of 736 participants on this trial were later examined
by serial sectioning and IHC. Occult nodal metastases were
found in 7% by serial sectioning and in 20% by IHC. These
metastases, detected by either method, were associated with a
higher risk of recurrence.
A prospective evaluation of the survival impact of IHC
metastases has been reported by Hansen et al. [19]. The SN
of 696 patients were examined by hematoxylin and eosin
(H&E) and IHC. The patients were divided into four groups:
Group I, SN-negative; Group II, SN IHC-positive/H&E neg-
ative or equivocal; Group III, SN micrometastases ≤2 mm;
and Group IV, SN H&E macrometastases >2 mm. At a
median follow-up of 38 months, the size of the SN metasta-
sis was a significant predictor of disease-free survival (DFS)
(p = 0.0001) but not overall survival (OS) (p = 0.0520).
There was no significant difference in DFS or OS between
the SN-negative and the SN IHC-positive patients The
authors concluded therefore that treatment decisions should
not be made on the basis of SN IHC positivity.
In conclusion, axillary node status is the most consis-
tent prognostic factor used in adjuvant therapy decision
making. It is standard practice to administer adjuvant ther-
apy to patients with lymph nodes that are positive using
H&E staining. There is, however, increasing use of SN
biopsies to stage the axilla. Patients with lymph nodes that
are positive using H&E staining are offered adjuvant ther-
apy. Therapy for patients that have SN positive by IHC
only is a more complex decision, and other factors, such as
tumor size, grade, hormone receptor status, and age become
more influential.
Tumor Size
Tumor size correlates with the presence and number of
involved axillary lymph nodes and is also an independent
prognostic factor, with distant recurrence rates increasing
with larger tumor size. The SEER database includes 13,464
women with node-negative breast cancer. Patients with
tumors <1 cm had a 5-year OS of close to 99% compared with
89% for tumors between 1 cm and 3 cm and 86% for tumors
between 3 cm and 5 cm [20]. This association persists with
longer follow-up. Rosen et al. examined the relationship
between tumor size and 20-year recurrence-free survival and
found a significant association, with a 20-year recurrence-free
survival of 88% for tumors ≤1 cm, 72% for tumors 1.1 cm to
3 cm, and 59% for tumors between 3.1 cm and 5 cm [21].
Furthermore, median time to the development of metastatic
disease also shortens as tumor size increases [20-23].
For node-negative patients, tumor size is the most pow-
erful prognostic factor and is routinely used to make adju-
vant treatment decisions. In general, patients with a tumor
size of >1-2 cm warrant consideration of adjuvant therapy
since they may have a distant recurrence risk of ≥20%.
Tumor Type/Grade
The pathologic characteristics of the tumor have prog-
nostic significance. Certain subtypes such as tubular, muci-
nous, and medullary have a more favorable prognosis than
unspecified breast cancer [24-26].
In an attempt to improve interobserver variability, mul-
tiple grading systems have been proposed, with the most
widely accepted being the Scarff-Bloom-Richardson (SBR)
classification [27]. Mitotic index, differentiation, and pleo-
morphism are scored from 1 to 3 and the scores from each
category are totaled. Tumors with scores from 3 to 5 are
well differentiated (grade 1), from 6 to 7 are moderately
differentiated (grade 2), and 8 to 9 are poorly differentiated
(grade 3). A correlation between histologic grade as deter-
mined by SBR and 5-year DFS has been demonstrated in a
study of 1,262 women [28]. Patients with an SBR score of
3 had a relative risk of recurrence of 4.4 compared with
those with an SBR of 1. In conclusion, tumor grade does
have prognostic significance and is primarily used to make
decisions for lymph node-negative patients with borderline
tumor sizes.
Lymphatic and Vascular Invasion
Peritumoral lymphatic vessel and vascular invasion
(LVI) has been demonstrated to have prognostic signifi-
cance for the risk of local and distant recurrence. At 20
years of follow-up, Rosen et al. noted a correlation between
lymphovascular invasion and the risk of recurrence and
death [29]. The recurrence rate for women with LVI-posi-
tive stage I disease was 38% compared with 22% for those
with LVI-negative disease. In addition, the International
Breast Cancer Study Group randomized 1,275 women with
by on October 17, 2009
www.TheOncologist.com
Downloaded from

Page 4

Cianfrocca, Goldstein
609
node-negative breast cancer to a single cycle of periopera-
tive chemotherapy or no systemic adjuvant therapy and
demonstrated that the presence of LVI was associated with
a 15% increase in the 5-year recurrence risk, and this effect
was independent of whether or not they received adjuvant
therapy [30]. Lymphatic and vascular invasion does have
prognostic significance and is primarily used to make deci-
sions for lymph node-negative patients with borderline
tumor sizes.
Proliferation Markers
Various methods of measuring the proliferative rate of
tumors have been evaluated in an attempt to correlate them
with prognosis. These putative markers include the S-phase
fraction (SPF), thymidine labeling index, mitotic index, and
IHC analyses using antibodies directed against proliferation
antigens such as Ki-67 and proliferating-cell nuclear anti-
gen [31-33]. Many studies are limited by a lack of consis-
tent methodology as well as a lack of information regarding
systemic therapy and other prognostic variables. Several
trials, however, have demonstrated an association between
SPF and prognosis. After adjusting for tumor size, lymph
node status, ploidy, age, and adjuvant systemic therapy,
analysis of the San Antonio Breast Cancer Data Base
revealed that a high SPF was associated with a relative risk
of death of 1.29 (p < 0.0001) [34]. Similar results were
observed on the NSABP B-14 trial in which women with
ER-positive, node-negative tumors were randomized to
receive 5 years of tamoxifen or placebo [35]. After adjust-
ing for tumor size, age, and treatment, patients with high
SPF tumors had a higher risk of both recurrence and death
compared with those with low SPF tumors. Furthermore, an
intergroup trial evaluated the natural history of 1,208 node-
negative patients who were determined to be low risk based
on tumor size, hormone receptor status, and SPF and were
followed without adjuvant therapy. Low-risk patients either
had hormone receptor-positive tumors that were <2 cm
with a low SPF or had tumors that were too small for recep-
tor analysis. High-risk patients had a tumor size ≥2 cm or
tumors that were ER and PR negative. Low-risk patients
were observed without adjuvant systemic therapy, while
high-risk patients were randomized to receive CMF or
cyclophosphamide, adriamycin, and 5-flourouracil (CAF).
At 5 years of follow-up, the low-risk women had a DFS of
88%-89% and an OS of 96%-97%, regardless of whether
they were classified as being low risk by tumor size alone or
by a low SPF [36]. This finding suggests that a low SPF may
be used in conjunction with other prognostic factors to iden-
tify a group of node-negative patients that are sufficiently
low risk to not warrant a recommendation of adjuvant
systemic therapy.
In conclusion, proliferation factors, such as SPF, do
have prognostic significance. An elevated SPF is primarily
used as justification to administer adjuvant therapy to lymph
node-negative patients with borderline tumor sizes. A low
SPF, however, may be used to identify a group of lymph
node-negative patients who may not require adjuvant therapy.
Ethnicity and Patient Age at Diagnosis
African American and Hispanic women have a
decreased survival from breast cancer compared with white
women [37-39]. The source of this disparity is likely multi-
factorial, including issues such as lack of access to care
resulting in a higher stage at diagnosis. There are data, how-
ever, to suggest that survival may be worse for African
American women, even after adjusting for disease stage [40].
Many studies evaluating the influence of age on out-
come in breast cancer have been small and have had con-
flicting results [41-45]. Two relatively large trials have,
however, demonstrated a worse prognosis for patients
younger than 35 years of age, even after adjustment for other
prognostic factors [46, 47].
Ethnicity and age at diagnosis may be used to identify
a group of patients who have a higher risk recurrence. They
should be used, however, as an adjunct to other prognostic
factors that are better validated such as tumor size.
PROGNOSTIC AND PREDICTIVE FACTORS
ER/PR Status
The presence of estrogen and progesterone receptors in an
invasive breast carcinoma is both prognostic and predictive.
Its prognostic effect is difficult to evaluate in that it must be
assessed in the absence of adjuvant tamoxifen. The NSABP
B-06 trial randomized women with early-stage breast cancer
to mastectomy, lumpectomy alone, or lumpectomy followed
by radiation therapy [48]. No adjuvant systemic therapy was
administered. The women with ER-positive tumors had a
5-year DFS of 74% and OS of 92%, while the women with
ER-negative tumors had a 5-year DFS and OS of 66% and
82%, respectively.
Studies with longer follow-up, however, suggest that the
prognostic significance of hormone receptors may not per-
sist long-term. Hilsenbeck et al. demonstrated an improved
prognosis for ER-positive tumors during the first 3 years of
follow-up but not after 3 years [49]. It is possible that the
presence of estrogen or progesterone receptors merely pre-
dicts for a more indolent, slower growing tumor with longer
times to disease recurrence.
The presence of estrogen or progesterone receptors is,
however, a powerful predictive factor for the likelihood of
benefit from adjuvant tamoxifen. The most recent Early Breast
by on October 17, 2009
www.TheOncologist.com
Downloaded from

Page 5

610
Prognostic and Predictive Factors in Early-Stage Breast Cancer
Cancer Trialists’ Collaborative Group update of randomized
trials using adjuvant tamoxifen, published in 1998, included
37,000 women [4]. These data demonstrated that 5 years of
adjuvant tamoxifen led to proportional reductions in the risk
of recurrence and mortality of 47% and 26%, respectively, of
patients with ER-positive tumors. This proportional reduc-
tion in mortality was similar for node-negative and node-pos-
itive patients. This translated to absolute mortality reductions
of 5.6% for those with node-negative disease and 10.9% for
those with node-positive disease. Five years of adjuvant
tamoxifen also led to a proportional reduction of 47% in the
risk of contralateral breast cancer, which was the rationale for
the NSABP P-1 tamoxifen prevention trial [50]. There was
no benefit for tamoxifen in hormone receptor-negative
women [4].
The prognostic significance of estrogen or progesterone
receptors is limited. Its optimal use is as a predictive factor
for the benefit of adjuvant tamoxifen therapy. As a result of
the data discussed above, all hormone receptor-positive
women who warrant adjuvant systemic therapy should
receive hormonal therapy unless otherwise contraindicated.
HER2/neu
The c-erbB-2 (HER2/neu) proto-oncogene is located on
17q21 and encodes an Mr 185,000 transmembrane glyco-
protein, p185HER2, with intrinsic tyrosine kinase activity
homologous to the epidermal growth factor receptor [51]. It
is amplified and/or overexpressed in approximately 30% of
human breast tumors [52]. Overexpression is associated
with increased tumor aggressiveness, increased rates of
recurrence, and increased mortality in node-positive
patients, while the influence in node-negative patients is
more variable [53-56]. Interpretation of many studies of
HER2, however, is limited by variability in the methods
used to detect overexpression, definition of positivity, and
that most are retrospective subset analyses.
Retrospective studies have suggested that HER2/neu
overexpression may also have a predictive role for response
to chemotherapy and endocrine therapy. A Southwest
Oncology Group study, for example, randomized patients
to tamoxifen alone or tamoxifen plus CAF and found that
CAF improved the outcomes of the HER2/neu-positive
women [57].
Several studies have suggested that HER2/neu overex-
pression may be associated with resistance to alkylator-
based chemotherapy (Table 1). The International (Ludwig)
Breast Cancer Study Group Trial V randomized node-neg-
ative patients to receive either no adjuvant therapy or a sin-
gle cycle of perioperative chemotherapy with CMF, and
node-positive patients to either prolonged chemotherapy (a
perioperative cycle and six postoperative cycles of CMF) or
a single perioperative cycle. HER2/neu overexpression was
seen in 16% of the node-negative patients and 19% of the
node-positive patients. For both node-positive and node-
negative patients, the benefit of chemotherapy was greater
for the HER2/neu-negative patients [58]. Similarly, Allred
et al. demonstrated that HER2/neu-negative patients who
received adjuvant CMF plus prednisone had an improved
DFS compared with those with HER2/neu-positive tumors
[59]. Miles et al. examined the relationship between
HER2/neu status and outcome in 274 node-positive women
who were randomized to receive six cycles of adjuvant CMF
or no adjuvant therapy [60]. Although all of the treated
women appeared to benefit from adjuvant CMF, the
improvement in survival was less in the HER2/neu-positive
patients. Results from the first CMF randomized trial with a
20-year follow-up, however, do not support a negative rela-
tionship between HER2/neu overexpression and response to
adjuvant CMF [61].
HER2/neu expression may also predict benefit from
adjuvant anthracyclines (Table 2). A retrospective analysis
of 141 breast tumors included in a multicenter randomized,
phase III trial comparing adjuvant CMF to 5-fluorouracil,
doxorubicin, cyclophosphamide (FAC) was performed,
demonstrating that the survival for the FAC arm was simi-
lar for HER2/neu-positive and HER2/neu-negative patients
(p = 0.1) [62]. In contrast, the survival for the CMF arm was
significantly worse for HER2/neu-positive compared with
HER2/neu-negative patients (p = 0.006), supporting the
concept that anthracycline-based chemotherapy offers a
Table 1. HER2/neu overexpression and resistance to adjuvant CMF
Author
Gusterson et al. [58]
Trial
IBCSG
Design
LN–: PeCT versus Nil
LN+: PeCT versus CMFP
CMF versus Nil
CMF versus Nil
n of ER+patients
1,506
Does HER2 predict resistance?
Maybe
Allred et al. [59]
Menard et al. [61]
Intergroup 001
Istituto Nazionale Tumori
306
337
Maybe
No
Abbreviations: IBCSG = International Breast Cancer Study Group; LN = lymph node; PeCT = perioperative chemotherapy; CMFP = CMF plus
prednisone
by on October 17, 2009
www.TheOncologist.com
Downloaded from