Nguyên Lý Và Kỹ Thuật Trong Phẫu Thuật Ung Thư Vú Tạo Hình

Neoadjuvant Therapy
31
3.3 Staging
3.3.1 Current Staging Guidelines
Patients who are candidates for NCT are those with larger tumors or
with clinically palpable lymph nodes; therefore it is important to rule out
metastatic disease prior to initiation of treatment. The current NCCN guide-
lines recommend a metastatic workup for patients with stage IIb cancers or
higher, similar to the group of patients who would be considered for NCT.9
NCCN staging guidelines consist of blood work,a chestX-ray,a computerized
axial tomography (CAT) scan of the abdomen and pelvis, and a bone scan.
Any abnormalities should be further evaluated with subsequent imaging or a
positron emission tomography/CAT scan. For example, an abnormality on a
bone scan may lead to magnetic resonance imaging (MRI) of the suspicious
lesion or an 18F-fluorodeoxyglucose-positron emission tomography (FDG-
PET) scan to assess the lesion for metabolic activity. It is also important to
obtain pathologic confirmation of metastatic disease because this will render
the patient stage IV and alter the treatment plan.
3.3.2 The Role of Imaging
The ultimate goal of NCT is to eradicate the tumor. Often this means the
tumor’s appearance after chemotherapy is drastically different compared
with its appearance prior to initiation of treatment, both radiographically
and clinically. This highlights the importance of thoroughly evaluating
malignant lesions with imaging studies not only prior to the initiation
of chemotherapy but also prior to surgical treatment. This also reinforces
the need to place a tumor marker prior to the initiation of chemotherapy.10
There are instances in which the patient has a complete radiologic response
and the tumor marker is needed in order to proceed with breast conserva-
tion with needle localization.
3.3.3 Mammogram
The importance of assessing the extent of the tumor prior to NCT is
paramount. Often this workup begins with a clinical breast examina-
tion and standard imaging modalities, consisting of mammography and
sonography. Mammography is usually the initial imaging modality used to
evaluate the breast, either before or after diagnosis of carcinoma. While
mammography has been shown to be more accurate in patients with less-
dense parenchyma, it can also be helpful in detecting calcifications in or
surrounding a lesion. However, mammograms are more limited when eval-
uating tumors located in the far posterior breast against the chest wall.11

32
Principles and Techniques in Oncoplastic Breast Cancer Surgery
Mammography also incompletely visualizes nodal basins other than the
low axilla. In 2 small studies, mammography alone or in combination with
clinical breast exam detected invasive breast cancers in about 81% to 86%
of patients with known disease.12,13 As an adjunct to mammography, the
ultrasound examination of the breast can be useful in the context of a more
complete assessment of the breast prior to treatment.12
It has been our observation that microcalcifications do not resolve with
NCT in spite of good pathologic response. All suspicious residual microcal-
cifications seen around the tumor bed should be completely excised during
surgical intervention. Hence patients who present with extensive microcal-
cifications that cannot be excised by breast conservation and an oncoplastic
procedure should have a mastectomy.
3.3.4 Ultrasonography
In the early neoadjuvant trials, a physical exam with calipers and ultra-
sound measurement of the tumor was standard. In the same trials already
listed, diagnostic ultrasonography detected 94% to 97% of invasive breast
cancers, and detected multifocal or multicentric disease in a total of 14% of
patients.12,13 For patients who are being evaluated for breast conservation,
it is important to obtain tissue diagnosis of all suspected lesions prior to
initiation of NCT. This will allow for an evaluation after chemotherapy to
assess treatment response, as well as direct surgical planning. It may also
exclude patients with multicentric disease for breast conservation.
Mammography and sonography may be used to not only perform a
percutaneous needle biopsy to confirm the presence of the tumor but also
to leave a metal-marking clip in the area of biopsy.10 Some physicians prefer
to leave a clip at the center of the tumor, while others prefer to mark the
areas of tumor which are furthest apart, which may have implications for
the volume of surgical resection needed. We have relied on a clip in the
center of the tumor as we limit our resection to the residual disease, and
we do not necessarily excise all of the original tumor bed if the patient had
achieved a pathologic response to chemotherapy. Confirmation of proper
clip placement should be obtained with a postplacement mammogram,
even if the clip is placed by ultrasound guidance. Again, these markers are
of the utmost importance in the case of a complete radiologic response. In
this situation, the clips are the only evidence of the location of the prior
tumor. If breast conservation is performed, these clips can then be used
for needle localization at the time of definitive surgery. In patients with
multiple lesions, differently shaped metallic clips should be placed in each
lesion to help direct which area to excise after completion of chemotherapy
if breast conservation becomes an option.

Neoadjuvant Therapy
33
Ultrasonography can be helpful not only in measurement of primary
breast tumors but also in assessment of nodal basins. The appearance of a
lymph node with a change in architecture consisting of a thickened cortex or
loss of the fatty hilum leads to the suspicion of regional metastatic disease.
These characteristics of macrometastatic disease are easily seen by utilizing
current ultrasound technology. Ultrasonography has difficulty visualizing
micrometastases, and therefore pathologic evaluation is still necessary.11,14
Krishnamurthy et al. reported results comparing ultrasound-guided fine
needle aspiration (FNA) of suspicious axillary lymph nodes with those
excised at complete axillary node dissection.14 FNA correctly identified
positive lymph nodes in all patients who had 3 or more positive lymph
nodes. It also correctly identified positive lymph nodes in 93% of patients
who had a metastatic deposit > 0.5 mm. The sensitivity of the ultrasound-
guided FNA of the axillary lymph node was 86%, and the specificity was
100%. The positive predictive value was 100%, and the negative predictive
value was 67%, with an overall diagnostic accuracy of 79%. Ultrasonogra-
phy can be used to evaluate not only the axillary lymph nodes but also the
internal mammary and supraclavicular nodes.
Although not performed routinely, targeted ultrasonography can be
used as an adjunct to an abnormal clinical finding or a radiologic abnor-
mality after NCT. A change in the appearance of lymph nodes after
chemotherapy might indicate a response to treatment and can be valuable
prognostically.15 The information obtained from an axillary ultrasound
can be invaluable. This aside, pathologic evaluation of the lymph node
needs to be performed during treatment, whether it is prior to or after
NCT, for ultimate surgical and radiation treatment planning.
3.3.5 Magnetic Resonance Imaging (MRI)
Functional imaging has emerged as a helpful tool in the assessment of
breast cancer patients eligible for NCT. For newly diagnosed breast cancer
patients, MRI has a sensitivity > 90%.16 However, more importantly, up
to 10% of patients with a newly diagnosed breast cancer have an occult
lesion detected by MRI in either the ipsilateral or contralateral breast.16
Finally, MRI is superior to mammography at visualizing far posterior
lesions, including those with chest-wall involvement.11
In patients being considered for NCT, it is important to biopsy any
suspicious lesions detected by MRI. Biopsy can be performed via MRI if
necessary, but also by targeted ultrasonography or mammography if the
lesion is seen on these studies subsequent to the diagnostic MRI. This
allows for adequate stratification of patients who remain eligible for NCT
and identifies lesions that need to be excised at definitive surgery. It is

34
Principles and Techniques in Oncoplastic Breast Cancer Surgery
important to remember to leave a clip in any area where an image-guided
biopsy is performed.10
Some studies have shown that MRI is more accurate at predicting resid-
ual disease after NCT than mammography or ultrasonography.17,18 In fact,
it has been suggested that a pCR can be predicted by lack of residual
enhancement on a postneoadjuvant MRI. This is an advantage of func-
tional imaging in that an MRI not only measures the static size of the
tumor but also the viability of the tumor cells by using contrast-enhancing
kinetics.19 For example, after NCT, some masses are still present on phys-
ical exam and standard imaging. On MRI, however, some of these tumors
no longer take up gadolinium in a suspicious pattern, indicating these cells
are likely necrotic from the chemotherapy. Rather than relying on a clin-
ical exam in which a palpable mass is still present after NCT, an MRI,
revealing a decrease in suspicious uptake, may more confidently allow for
a recommendation for an attempt at breast conservation and decrease the
surgical resection volume. Performing an MRI prior to and after NCT can
be a helpful adjunct to assess tumor response and has become an inte-
gral part of the workup of the patient undergoing NCT.20 It has been
our personal experience, however, that lack of uptake of gadolinium may
be related to partial disruption of vascular supply of the tumor without
complete pathologic response.
One of the most important aspects with respect to imaging in the neoad-
juvant patient is to repeat the studies after treatment. Using a combination
of modalities — mammography, ultrasonography, MRI, and clinical exam-
ination — increases the sensitivity and more accurately measures residual
tumor size. This allows for both assessment of tumor response and better
surgical planning.11 The surgeon has to be cognizant that if MRI is used
as the sole measure of tumor size, this may lead to more extensive surgery.
Houssami et al. reported MRI led to more extensive surgery in 11% of cases
and resulted in unnecessary conversion to mastectomy in between 3% and
33% of cases (average: 8%).21
3.4 Predictors of Response
It is helpful to know which patients will have an appropriate response to
chemotherapy in order to help stratify those patients who will be eligible
for breast conservation. This allows the surgeon and the medical oncol-
ogist to set appropriate expectations. In addition, it allows for planning
of chemotherapy regimens according to tumor biology. Overall, there is a
wide range in the response to NCT. The measurements of response include
both clinical and pathologic response. In general, patients with triple neg-
ative and HER2 positive biologic profiles have a higher overall response

Neoadjuvant Therapy
35
rate, both clinically and pathologically. Tumors that are hormone receptor-
positive have lower clinical and pathologic response rates.
NCT may cause cancer cells to respond in numerous ways that affect
the subsequent size and shape of the residual tumor. Some tumors have a
response in the majority of the cells that causes them to shrink in a concen-
tric pattern. These tumors exhibit a decrease in their overall size, possibly
making them eligible for breast conservation after treatment. This is in
contrast to other tumors that do not achieve the same reliable pattern of
response. These tumors may have a heterogeneous response to treatment,
which leaves islands of treated cells intermingled with viable cancer cells.
Unfortunately, these tumors still retain most of their original overall size
and may contribute to the failure of breast conservation in this popula-
tion. Examining the different biological subtypes of breast cancer sheds
light on the response to treatment and the likelihood of achieving breast
conservation.
3.4.1 Response to NCT Based on Molecular Subtype
3.4.1.1 Triple negative
Approximately 15% of all breast cancers are estrogen and progesterone
receptor negative and HER2 receptor negative, otherwise known as triple-
negative breast cancer (TNBC). As a group, these tumors tend to be
more aggressive with poorer outcomes. One of the traditional choices of
chemotherapy for TNBC has included anthracycline-based regimes and
taxanes. Recent studies have also examined the role of platinum-based
chemotherapy in the patient with TNBC. There are also a number of
emerging targets in this particular population, including the poly-ADP-
ribose-polymerase (PARP) inhibitors, which have recently been included
in select neoadjuvant trials. It appears from numerous studies that patients
with TNBC have the highest clinical and pathologic response to NCT. This
fact increases the number of patients potentially eligible for breast conser-
vation after NCT.
In the largest series published to date, Liedtke et al. described the
pCR rate in their series of 225 TNBC patients who received NCT. This
group was compared to 863 patients not having TNBC who also underwent
NCT. In this study, 22% of the TNBC patients achieved a pCR compared
with 11% in non-TNBC patients (p = 0.034).22 This is in keeping with
previously reported numbers in smaller studies in which the range of pCR
ranged from 27% to 38%.23,24 The authors of this study noted an increased
pCR rate in patients treated solely with different anthracycline-based reg-
imens with or without a prior taxane. In this study by Liedtke et al., there

36
Principles and Techniques in Oncoplastic Breast Cancer Surgery
was not a significant pCR for a single agent-based taxane regimen, but
only a trend.
In addition to tumor response, this study also reported final surgi-
cal treatment in the TNBC versus non-TNBC groups. Those patients
with TNBC and a pCR, or 22% of the group, should have been eligible
for BCS after NCT. Encouragingly, this number was higher than 22%,
with 40% of TNBC patients undergoing BCS compared with 30% of non-
TNBC patients. This would indicate that even patients with only partial
responses were able to achieve breast conservation, perhaps with larger
surgical volume.
Platinum-based chemotherapy regimens are another group of agents
that have been investigated in the setting of TNBCs. Platinum-based ther-
apies have been utilized in this population as second-line treatment or for
metastatic disease for some time, but only recently have they been utilized
for a few small neoadjuvant trials. In one such trial, 22 patients with TNBC
were treated with cisplatin monotherapy, and 5 (23%) of the patients had
a pCR.25 In another small study, a group of BRCA-1 mutation carriers
with TNBC achieved a 72% pCR.26 However, the surgical outcome was
not reported, but theoretically, all of these patients would be eligible for
BCS. Given this information, platinum-based agents may hold promise in
some patients with TNBC and allow them to achieve breast conservation.
A list of pCR rates based on different chemotherapy regimens for TNBC
can be found in Table 3.1.
3.4.1.2 HER2 positive
Another group of locally advanced breast cancer (LABC) patients in
whom NCT holds promise for BCS are those with HER2 positive tumors.
Table 3.1.
Pathologic complete response rates in triple-negative breast cancer
(TNBC) patients in the neoadjuvant setting.62
Study
Chemotherapy regimen
pCR rate
Dana Farber
Cisplatin
22%
MGH/Dana Farber/BID
Cisplatin +/−bevacizumab
16%
Pomeranian Medical University
(BRCA1 + patients)
Cisplatin
72%
University of Texas
MD Anderson Cancer Center
Retrospective various regimens
22%
University of Texas
MD Anderson Cancer Center
Retrospective paclitaxel
followed by FACa
45%
a5-fluorouracil, doxorubicin, cyclophosphamide (FAC). pCR, pathologic complete
response.

Neoadjuvant Therapy
37
Table 3.2.
Pathologic clinical response to neoadjuvant chemotherapy
with or without trastuzumab for HER2 overexpressing tumors.
Study
With trastuzumab
Without trastuzumab
Buzdar27
65%
26%
Paluch-Shimon29
75%
24%
Peintinger30
73%
34%
Multiple randomized trials have now been published, demonstrating a
higher rate of pCR rates to NCT with the addition of trastuzumab for
HER2-overexpressing tumors (Table 3.2).27−30
In 2005, data from Buzdar et al.27 demonstrated a 62.5% pCR rate
with the addition of trastuzumab to a neoadjuvant protocol of four cycles
of paclitaxel plus fluorouracil (5FU), epirubicin, and cyclophosphamide
(FEC), compared to a 26% pCR with conventional chemotherapy in the
neoadjuvant setting, without increase in clinical congestive heart failure.
This was irrespective of hormone status, although patients with hormone-
positive tumors received endocrine treatment for 5 years in the adjuvant
setting. Although the trial was small and only enrolled 34 patients, it
was closed early due to the superiority of the trastuzumab arm. Subse-
quent trials, both prospective and retrospective, have demonstrated similar
findings.27
This same group then published data in which they examined the effect
of hormone-receptor status on the pCR rate.30 The trial included the poten-
tial addition of trastuzumab with various combinations of standard NCT.
There were multiple variables, including dosing schedules of taxol weekly
or once every 3 weeks, followed by 5-fluorouracil, doxorubicin or epirubicin,
and cyclophosphamide. The results demonstrated that patients with hor-
mone receptor positive-tumors had a significantly higher pCR with the
addition of trastuzumab. The highest pCR rate of 73% was achieved in the
group receiving trastuzumab plus 3-weekly taxol followed by FEC. These
findings were similar to the pCR rate seen in hormone receptor-negative
patients who also received trastuzumab with the neoadjuvant protocol.30
Early in 2010, the results from the neoadjuvant herceptin (NOAH) trial
were released in which patients with locally advanced or inflammatory
breast cancer were given either NCT with trastuzumab followed by adju-
vant trastuzumab, or chemotherapy alone.28 This trial also included an
HER2 negative cohort treated with the same chemotherapy regimen, which
consisted of doxorubicin, paclitaxel, cyclophosphamide, methotrexate, and
5-fluoruracil. Results of this trial demonstrated improved event-free sur-
vival, OS, and clinical and pathologic tumor responses.

38
Principles and Techniques in Oncoplastic Breast Cancer Surgery
Remembering that a pCR leads to a higher rate of breast conservation,
the addition of trastuzumab to NCT regimens for patients with HER2
positive tumors naturally suggests a higher rate of BCS in this group of
patients.
3.5 Gene Profiling
Some
researchers
have
evaluated
the
role
of
the
70-gene
profile
(Mammaprint®) in predicting response. Straver et al.31 categorized the
tumors of 167 patients by using the 70-gene profile and identified
23 patients with good prognosis signature and 144 with poor prognosis
signature or poor prognosis. All patients had received NCT. None of the
good signature prognosis patients achieved a pCR in the breast and axilla,
whereas 20% in the poor prognosis group had no residual tumor identi-
fied in the breast and axilla (p = 0.015). When looking for pCR or near
pCR in the breast, only 2 of 14 (9%) in the good prognosis group compared
to 53 of 144 (37%) in the poor prognosis group (p = 0.008) achieved that
end point.
3.6 Hormone Receptor-Positive Tumors
Multiple studies have demonstrated a reduced pCR to NCT for hormone
receptor-positive tumors.32 With this observation, the question became
whether a better method existed to reduce the size of locally advanced
breast cancers in the hormone receptor-positive population to facilitate
breast conservation. This led to the undertaking of trials that examined
the effect of neoadjuvant hormonal therapy with either tamoxifen or an aro-
matase inhibitor (AI) on these tumors. While trials examining the effect
of NCT in the TNBC and HER2 positive breast cancer patient tend to
report clinical and pathologic response rates, which can be used as a surro-
gate for potential for BCS, pCR rates for tumors treated with neoadjuvant
hormonal therapy tend to be low, in the order of approximately 2% to
3%.32 Rather, these trials report clinical response rates and percentage of
patients undergoing BCS, which is much more in keeping with the focus of
this chapter (Table 3.3).4,33,34
For premenopausal women, the choice for neoadjuvant hormonal ther-
apy is tamoxifen or ovarian suppression with an AI, much like adjuvant
treatment. However, there is little data available for neoadjuvant tamox-
ifen in the premenopausal patient. This is because premenopausal women
with LABC are more likely to be offered chemotherapy in a neoadjuvant
setting rather than hormonal treatment. The postmenopausal patient with

Neoadjuvant Therapy
39
Table 3.3.
Pathologic clinical response rate to neoadjuvant
chemotherapy, based on tumor hormone status.
Estrogen receptor-
Estrogen receptor-
Study
positive
negative
Buzdar, 200333
5%
21%
Tan, 200934
10%
54%
Rastogi, 20084
6%
14%
breast cancer has been more rigorously studied in the neoadjuvant hor-
monal setting. In this population, the reported breast conservation rate for
neoadjuvant tamoxifen is in the range of 30% to 35%.35,36 However, since
aromatase inhibitors are the first line of treatment for postmenopausal
women with hormone receptor-positive breast cancer, the main focus of
research has been on this group of medications.
The results from the first clinical trial comparing tamoxifen to an AI,
anastrozole or tamoxifen alone, or in a combination (ATAC) trial, showed
that anastrozole was superior to tamoxifen alone or with both drugs in
combination with respect to DFS in the adjuvant setting.37 Based on these
findings, the immediate preoperative anastrozole, tamoxifen, or combined
with tamoxifen (IMPACT) trial, a multicenter, double-blind randomized
trial that compared anastrozole alone, tamoxifen alone, or both drugs in
combination, was developed in the neoadjuvant setting.36 Patients were
given the determined medication for 12 weeks prior to surgery. The primary
endpoint was clinical overall response rate. Although, there was no statisti-
cally significant difference in overall response between patients treated with
anastrozole or tamoxifen alone, there was a trend. In subgroup analyses
of patients who were deemed to need a mastectomy prior to treatment,
there was a significantly different number of patients who were down-
graded from mastectomy to breast conservation in both medication groups
(44% versus 22%, p = 0.03). However, when subset analyses examined
the surgical outcome in each medication group, 45% of patients initially
requiring mastectomy were downgraded to breast conservation after neoad-
juvant anastrozole compared to 36% of patients after tamoxifen. This trial
concluded that neoadjuvant anastrozole was more effective at downstaging
patients than tamoxifen.
Similarly, the preoperative arimidex compared to tamoxifen (PROACT)
trial sought to compare each single agent with or without chemotherapy
in the neoadjuvant setting with respect to overall response measured by
clinical exam and ultrasonography, as well as rate of downgrading to breast
conservation.38 This large, randomized, double-blind, double-dummy trial
found that anastrozole was at least as effective as tamoxifen in obtaining an

40
Principles and Techniques in Oncoplastic Breast Cancer Surgery
overall response and breast conservation rate, but not statistically different.
Together, the results of the IMPACT and PROACT trials showed that
anastrozole was as effective or better then tamoxifen as a single agent in
the neoadjuvant setting in postmenopausal women at achieving a clinical
response and downstaging to BCS.36,38
The use of neoadjuvant letrozole was studied in the P024 trial, another
large, multinational, double-blind trial in which letrozole alone versus
tamoxifen alone was given for 4 months prior to surgery in hormone
receptor-positive patients who were not eligible for BCS.32 The primary
endpoints were clinical and pathologic response rate determined by clin-
ical breast exam, with the secondary endpoint being the percentage of
patients who underwent BCS. The overall response to letrozole was found
to be statistically significantly higher than to tamoxifen (55% versus 36%,
p < 0.001). In addition, significantly more patients receiving neoadjuvant
letrozole were able to achieve breast conservation compared with those
taking tamoxifen (45% versus 35%, p = 0.02). Interestingly, in this study,
patients who were HER2 positive were found to have a higher overall
response rate with letrozole compared to tamoxifen (p = 0.0004). The
overall response rate for patients with HER2 positive, estrogen-receptor
positive breast cancers taking letrozole was similar to that of patients with
HER2 negative tumors (71%). Letrozole was also found to be more effective
than tamoxifen in tumors that were marginally estrogen-receptor positive.
The conclusion from the P024 trial was that letrozole was more effective
than tamoxifen at inducing an overall response to treatment and downgrad-
ing patients requiring mastectomy to BCS. Further studies have confirmed
these findings.32
The final aromatase inhibitor, examestane, has been examined in the
neoadjuvant setting in smaller, phase II trials. Of the 221 total patients in
these trials, the overall response rate ranged from 34% to 73%, with 56%
to 76% of patients undergoing BCS (Table 3.4).32,36,39−42
Table 3.4.
Percentage of breast-conserving surgery (BCS) with neoad-
juvant hormonal therapy.
Hormone treatment
Study
BCS, %
Tamoxifen
IMPACT36
36
P02432
35
Anastrozole
IMPACT36
44
Letrozole
P02432
45
Examestane
Barnadas39
56
Mlineritsch40
76
Takei41
90
Tubiana-Hulin42
57

Neoadjuvant Therapy
41
Now that it has been shown that aromatase inhibitors are effective at
achieving a good clinical response with possible downstaging from mastec-
tomy to breast conservation, it is important to ascertain which patients are
likely to benefit from hormone therapy alone rather than in combination
with chemotherapy. In the adjuvant setting, there are genomic assays that
stratify patients into groups according to recurrence risk. Those tumors
with low recurrence risk do not benefit from the addition of chemotherapy
to hormonal treatment. Conversely, in those patients with tumors with a
high risk of recurrence, the addition of chemotherapy confers a benefit to
endocrine treatment (OncotypeDX®, MammaPrint®). It may be helpful to
use this information in a neoadjuvant setting to determine if a patient will
benefit from the addition of NCT in addition to neoadjuvant hormones.
One study in the literature examined whether a 21-gene assay could pre-
dict response to neoadjuvant treatment with tamoxifen or anastrozole.43
The recurrence scores were based upon pretreatment core biopsies. Only
half of the samples yielded enough RNA for examination. The results
showed that patients with low recurrence scores tended to have a better
response to neoadjuvant hormonal therapy than those with intermediate
or high recurrence scores, regardless of which hormonal therapy was used.
This was a small pilot study, but opens an interesting new area of study
with respect to tumor genomics and individualized neoadjuvant treatment.
3.7 Invasive Lobular Carcinoma
Patients with lobular carcinoma tend not to have as favorable a response
to NCT as those with ductal carcinoma.44 Perhaps it is the morphologic
features of lobular carcinoma and its tendency to have multiple islands
of tumor over distances which confer a poor response. It may also be
attributed to the fact that the vast majority of lobular tumors are strongly
hormone receptor-positive and HER2 receptor-negative. As previously
stated, the response of hormone receptor-positive tumors to NCT is not
as good as their estrogen receptor-negative counterparts. Therefore, for
a patient with invasive lobular carcinoma, one must adequately rule out
multicentricity and one should consider neoadjuvant hormonal therapy if
breast conservation is desired.45
3.8 Surgery After NCT
3.8.1 Surgical Decision Making
There are still a fair number of patients who undergo mastectomy after
NCT who on pathologic review are found to have a pCR. For example,

42
Principles and Techniques in Oncoplastic Breast Cancer Surgery
Buzdar reported a 50% breast conservation rate after NCT in a group of
patients in whom there was a 65% pCR rate.27 This lends itself to the ques-
tion of what factors contribute to surgical recommendations. The major
factors that lead to a final surgical decision seem to include persistence of
a mass on clinical exam, persistence of imaging abnormality, and patient
preference.5,46
Considering that patient preference for unilateral or bilateral mastec-
tomy is a factor that is not necessarily under a physician’s control, the
other contributing factors — presence of a mass or imaging abnormality —
should be addressed prior to surgical recommendation. Although the pres-
ence of a mass may indicate residual tumor cells, it may not represent
viable disease. As noted in the previous imaging section, mammography or
sonography may show residual mass or density, but MRI may not reveal
suspicious enhancement kinetics. After NCT, it is important to adequately
assess the amount of viable tumor before making surgical recommenda-
tions; an MRI is a valuable tool in this process.
3.8.2 Resection Volume and Margins
As with any other breast-conservation procedure, it is just as important to
obtain clear margins when performing surgery after NCT. It is well known
that tumor cells at the inked edge of a specimen carry a high risk of local
recurrence. Since there are variable response patterns to tumors under-
going NCT, careful evaluation of specimen margins is paramount. It is
also important to take into account the pattern of response histopatholog-
ically, and correlate this with preoperative imaging. Should a tumor have
a heterogeneous response, it is theoretically possible to dissect between
islands of tumor to have a seemingly clear margin. It behooves the sur-
geon and pathologist to review and report the pattern of tumor response
to treatment and then compare the tissue to preoperative MRI. Should
there be any doubt as to the response pattern, or if there was questionable
enhancement on MRI prior to surgery, either the MRI should be repeated
or re-excision should be performed.
The important concept to remember when operating on a patient after
NCT is that one is removing only the tumor that is present after chemother-
apy and not what was once there before the initiation of treatment. This
means that resection volume should be smaller after NCT. This is why it is
important to mark the center of a tumor with a metal marker clip prior to
initiation of treatment rather than the areas of the tumor that are farthest
apart.10 Should a surgeon remove the entire area between edges of the pre-
vious viable tumor, this negates the effect of NCT. Theoretically, if the
furthest edges of the mass have been marked prior to chemotherapy, as the

Neoadjuvant Therapy
43
mass regresses, the clips should move closer together. This also facilitates
a smaller resection volume.
For patients who do not achieve a pCR, breast conservation may still
be attempted. Studies have shown that using multiple localizing wires may
help achieve breast conservation for larger lesions with fewer re-excisions.47
In one such study, 77.1% of patients with larger breast cancer had successful
BCS by utilizing multiple localizing wires, with 28% requiring more than
one excision to achieve clear margins.48 Utilizing oncoplastic techniques in
larger volume resections can help maintain cosmesis in these cases with
larger resection volumes.
Finally, it is important to remove any residual calcifications seen on
imaging after completion of chemotherapy. This is because calcifications
may represent intraductal disease, which does not respond to chemother-
apy. A mammogram prior to surgery can aid in surgical planning for
patients with residual calcifications. Ultimately, the decision for breast
conservation, and the decision on the number of operations that should
be performed to attempt this, belongs in a discussion between patient and
surgeon, and plays a role in the individualized care of each breast cancer
patient.5
3.8.3 Axillary Evaluation
In the surgical literature, the timing of axillary evaluation in the patient
receiving NCT is a current controversy. There are multiple studies sup-
porting axillary evaluation prior to chemotherapy as well as afterward.
3.8.4 Sentinel Node Biopsy Prior to NCT
It is accepted that the false-negative rate of sentinel node biopsy is < 5%.
Early studies which evaluated the sentinel lymph node after chemotherapy
reported false-negative rates of up to 10%, which some clinicians felt was
unacceptably high.49 In addition, radiation planning may be affected by the
timing of the sentinel node biopsy. If a sentinel node contains metastatic
disease prior to NCT, then a full axillary dissection is performed at the time
of definitive surgery, which may or may not lead to the diagnosis of addi-
tional metastatic nodes. Conventional radiation planning is based on 4 or
more involved lymph nodes. There are some patients, however, who might
be offered comprehensive radiation after breast conservation or postmas-
tectomy radiation with only 1 or more involved lymph nodes if other tumor
factors are present, such as lymphovascular invasion, extracapsular exten-
sion, or high mitotic index. If a metastatic sentinel node is not evaluated
prior to NCT and is then sterilized by the chemotherapy, the information

44
Principles and Techniques in Oncoplastic Breast Cancer Surgery
regarding that node is lost, and comprehensive or postmastectomy radia-
tion may not be offered. Recent meta-analyses involving patients with just
1 to 3 positive lymph nodes who did not receive postmastectomy radiation
had a poorer DFS and OS.50 In fact, many radiation oncologists prefer to
know the status of the lymph nodes prior to NCT in order to optimize the
radiation fields.
3.8.5 Sentinel Node Biopsy After NCT
There are also data to support the performing of a sentinel lymph node
biopsy after NCT. The sentinel node detection rate post-NCT has varied
from 89% to 97%. The false-negative rate has been reported to vary from
10% to 12%.49,51,52
The theoretical concept is the same for the axilla as it is for the breast;
why remove tissue that does not contain tumor? Just as a smaller volume
of resection in the breast is the norm in most patients undergoing NCT, an
axillary lymph node dissection in a patient without lymph node metastases
has no therapeutic value and subjects the patient to increased morbidity.
If the sentinel node is the only node involved with tumor in approximately
60% of patients, and the sentinel node is sterilized by chemotherapy, then
there should be no residual tumor in the axilla in the majority of patients,
obviating the need for a completion axillary dissection. Furthermore, NCT
has been shown to downstage the axilla in 20.0% to 37.5% of patients.31,53
Straver et al. identified 3 predictors of cPR in axillary nodes: triple
negative tumor; HER2-neu positive tumor; and when there is complete
pathologic response in the breast primary. By following the same concept
of NCT converting patients from mastectomy to BCT, one can extrapo-
late that by downstaging the axilla, 20.0% to 37.5% of the patients may
be spared an unnecessary axillary node dissection if the sentinel node is
performed after completion axillary node dissection.
A recent study published by Hunt et al. reported their experience with
timing of sentinel node surgery prior to or after NCT. In this large ret-
rospective trial from 1994 to 2007, there was no difference in the sentinel
node identification rate.54 More importantly, there was no difference in
local recurrence based on timing of the sentinel node biopsy. Criticisms
of this trial include the fact that only 15% of patients underwent sen-
tinel node biopsy after chemotherapy and that there was no reporting of
radiation fields, which also contributes to local control. Follow-up was not
addressed since this was a retrospective study.
A current trial by the American College of Surgeons Oncology Group
is designed to answer the questions of whether a sentinel node biopsy is

Neoadjuvant Therapy
45
reliable after chemotherapy and what percentage of the time the axilla is
sterilized by NCT. In this trial, patients undergoing NCT who are clini-
cally node negative will undergo axillary ultrasound, with aspiration of
any suspicious node prior to initiation of chemotherapy. If positive, the
patient then goes on to sentinel node biopsy and completion axillary dis-
section after completion of NCT. The results from this trial will be eagerly
anticipated.
Until further data are available, evaluation of the axilla must be per-
formed for the patient undergoing NCT. This may be done prior to or after
chemotherapy, depending on the decision of the multidisciplinary team and
the patient.
3.9 Locoregional Failure: Incidence and Predictors
There have been multiple series in the literature that have reported series
rates of local and regional failure post-NCT in patients who underwent
breast conservation (Table 3.5). The rate of local and regional recurrence
seems to have decreased over the years; this can be attributed to multiple
factors, including improvement in chemotherapy regimens, development of
targeted therapies, and perhaps better selection criteria. The 5-year rate of
locoregional recurrence (LRR) in the modern series over the past 12 years
has been around 8%, and the 10-year rate has been around 12%. Such rates
are quite encouraging and seem to be consistent among different centers.
Such relatively low recurrence rates are likely related to a selection bias as
Table 3.5.
Rates of local and regional failure post-neoadjuvant chemotherapy in
patients who underwent breast conservation.
% Locoregional recurrence BCS
Trial
Year
Follow-up, mos
5 years
10 years
Institut Bergonie63
1991
124
23
Institut Curie64
2001
93
16
21.5
NSABP B-18
1998
60
7.9
Milan CI65
1998
96
6.8 (8 years)
University of Texas
MD Anderson
Cancer Center66
2004
60
8.5
University of North
Carolina67
2002
70
10
Schwartz68
2006
75
7
12
BCS, breast-conserving surgery.

46
Principles and Techniques in Oncoplastic Breast Cancer Surgery
patients who undergo breast conservation are those who have achieved a
significant response to chemotherapy to permit breast conservation.
3.10 Predictors of Relapse
3.10.1 Tumor and Patient Characteristics
There are many patient and tumor factors that may contribute to relapse
after NCT. For example, Liedtke et al. reported a significantly decreased
PFS, a poorer PFS for patients who presented at a younger age, ductal
versus non-ductal histology, high nuclear grade, and high tumor stage.22
They also noted a decreased PFS for patients with TNBC compared with
non-TNBC (63% versus 76%). In fact, presence of the estrogen receptor or
progesterone receptor was associated with increased PFS, while the pres-
ence of the HER2 receptor showed a trend toward decreased PFS.
A prognostic index was developed and validated by a group at the Uni-
versity of Texas MD Anderson Cancer Center to predict LRR after NCT.
The scoring system relies on 4 predictors of ipsilateral and locoregional
tumor recurrence: Clinical N2 or N3 disease; residual pathologic tumor
size > 2 cm; a multifocal pattern of residual disease; and lymphovascular
invasion. The scoring system assigns a score of 0 (favorable) or 1 (unfavor-
able) to each criterion based on the presence or absence of each of these
factors. The final score is derived by adding the result from each individ-
ual variable with 5 possible final categories (0, 1, 2, 3, 4). In patients with
scores of 0, 1, or 2, the risk of LRR was the same for those who under-
went breast conservation or mastectomy. For patients with scores of 3 or
4, the risk of LRR was significantly higher for those who underwent breast
conservation when compared to mastectomy; however, the disease-specific
survival was similar. Such scoring may be helpful in selecting patients for
breast conservation after NCT.55,56
3.10.2 Residual Disease After NCT
The strongest predictor of relapse after NCT is failure of a pCR.57,58 It has
been shown that patients with residual disease after NCT have decreased
DFS and OS. When examining TNBC patients, those patients who do not
achieve a pCR have a decrease in OS, whereas TNBC patients who achieve
a pCR have a similar OS to non-TNBC patients.22,24 Similarly, patients
with overexpression of the HER2 receptor who did not achieve pCR were
found to have a higher likelihood of relapse.23
In the TNBC population, it has also been shown that the risk of relapse
and death is highest in the first 3 years after treatment. After that time

Neoadjuvant Therapy
47
period, the risk of relapse and death is similar to or even lower than non-
TNBC patients.22 Since current NCCN guidelines recommend follow-up
staging studies only for symptomatic patients, it becomes important to
carefully follow those patients treated with NCT, especially those with
residual disease. Surgical treatment of local failure should consist of a mas-
tectomy if the patient still has a breast, or excision of chest wall recurrence,
in combination with adjuvant regimens.
3.11 Future Directions
3.11.1 PARP Inhibitors
There are a number of novel therapies in phase II or III trials in the neoad-
juvant setting for the treatment of breast cancer. One such area of research
is the use of PARP inhibitors in the TNBC population. Both PARP
inhibitors and BRCA are both responsible for DNA damage repair. PARPs
are responsible for single-stranded DNA repair, and BRCA is responsible
for double-stranded DNA repair. In BRCA mutation carriers, when there is
DNA damage to the tumor cells that cannot be repaired by BRCA because
of a mutation, PARP repairs the DNA damage, thus allowing those tumor
cells to grow and thrive. PARP inhibitors prevent DNA repair in tumor
cells, leaving tumor cells without any mechanism for repair. This causes
tumor cells to enter an apoptotic pathway leading to their demise. The
use of PARP inhibitors BSI-021 and olaparib were the first 2 agents to
be investigated and reported. A study of 74 patients treated with BSI-021
at 5.6 mg/kg combined with carboplatin and gemcitabine versus the same
chemotherapeutic agents alone showed an increased objective response rate
(48% versus 15%), increased PFS (3.3 to 6.9 months), increased clinical
benefit (62% to 21%), and increased OS (5.7 to 9.2 months).59 Similarly,
in a phase II trial, the PARP inhibitor olaparib has been shown at a dose of
400 mg twice daily to improve PFS to 5.7 months with an overall response
rate of 41%.60,61 Currently, there are a number of ongoing clinical trials for
various PARP inhibitors for BRCA mutation carriers in the neoadjuvant
setting that hold exciting promise for this population with respect to local
surgical treatment and OS.
3.11.2 Vascular Endothelial Growth Factor
(VEGF) Inhibitors
Another area of ongoing research is the category of VEGF inhibitors. In
the neoadjuvant setting, there are a number of trials open to patients with
TNBC. Isakoffet al. used bevicizumab and cisplatin in the neoadjuvant

48
Principles and Techniques in Oncoplastic Breast Cancer Surgery
setting in a triple-negative population and achieved a 16% pCR rate.62
The Cancer and Leukemia Group B currently has an open trial, and the
University of Tennessee Cancer Institute has a planned trial to examine the
use of paclitaxil ± carboplatinum followed by AC ± bevicizumab in some
combination for TNBC patients in the neoadjuvant setting. These trials
should shed light on the question of whether a VEGF inhibitor improves
pCR, and more importantly, whether a VEGF inhibitor improves DFS
and OS.
While there are no trials currently open in the neoadjuvant setting
examining a VEGF inhibitor for HER2-positive breast cancer, the BETH
study (treatment of HER2-positive breast cancer with chemotherapy plus
trastuzumab versus chemotherapy plus trastuzumab plus bevacizumab)
will evaluate the efficacy of a VEGF inhibitor in the adjuvant setting. This
trial is sponsored by the NSABP; if results are positive, this could lead to
a neoadjuvant trial in the future.
In another area of research, there are alternative methods being eval-
uated using various predictive markers of pCR and survival. One such
trial, the I-SPY, is open to HER2 positive and TNBC patients. In this
trial, core biopsy specimens are obtained prior to chemotherapy, and var-
ious chemotherapeutic agents will be combined with novel therapies to
assess tumor response. Similarly, a study sponsored by the United States
Department of Defense is using core biopsy samples prior to NCT and
subjecting them to chemo assays to obtain potential therapeutic response.
This concept of personalizing breast cancer care based on tumor genomics
prior to initiation of treatment will certainly play a large role in the way
breast cancer is managed in the future. If more tumors can be oblit-
erated with chemotherapy prior to surgery, more patients will be eligi-
ble for breast conservation and the OS of our patients may be greatly
impacted.
3.12 Conclusion
The ultimate goal of NCT is to induce a complete pathologic response
in the primary breast cancer and any involved lymph nodes. Not only
does this render a patient eligible for BCS, but it also portents the best
overall outcome for the breast cancer patient. Currently, to achieve this, one
must consider the biological profile of the tumor and adjust chemotherapy
regimens accordingly. In the future, however, there may be novel techniques
which take into account the genomics of individual tumors, allowing a more
personalized approach to treatment.

Neoadjuvant Therapy
49
3.12.1 Patient Vignettes
3.12.1.1 Patient 1
A 36-year-old woman presented with a palpable mass in her lateral right
breast, which she detected on self-breast exam. The mammogram was neg-
ative for findings. A targeted ultrasound revealed a 3-cm solid mass in a
B cup breast [Figure 3.1(a)]. Core needle biopsy revealed a triple-negative
invasive ductal carcinoma. A clip was placed in the center of the mass at
the time of core needle biopsy. A bilateral breast MRI revealed no other
suspicious lesion in either breast [Figure 3.2(a)].
The patient desired breast conservation. Because she was going to
receive adjuvant chemotherapy, she elected to have her chemotherapy in a
neoadjuvant setting. This not only improved the likelihood of breast con-
servation but also improved her cosmetic outcome.
She underwent a sentinel node biopsy prior to start of chemotherapy,
which was negative for metastatic disease, and was then given docetaxel,
doxorubicin, and cyclophosphamide (TAC) 3-weekly for 6 cycles. She had
a cCR to treatment. After repeating her ultrasound and MRI to evalu-
ate response [Figures 3.1(b) and 3.2(b)], she underwent needle localized
lumpectomy followed by whole breast radiation. Figure 3.3 shows the
patient 6 months after completion of radiation treatment. Her pathol-
ogy revealed no further invasive carcinoma, but residual atypical ductal
(a)
(b)
Figure 3.1.
(a) A 3-cm solid mass in a B-cup breast revealed by targeted ultrasound.
(b) Repeated ultrasound to evaluate response.

50
Principles and Techniques in Oncoplastic Breast Cancer Surgery
(a)
(b)
Figure 3.2.
(a) No other suspicious lesion in either breast revealed by bilateral breast
MRI. (b) Repeated MRI to evaluate response.
(a)
(b)
Figure 3.3.
Six months after completion of radiation treatment.
hyperplasia. The size of her surgical specimen was 4.5 cm × 3.5 cm ×
1 cm. She remains disease free at 2 years.
3.12.1.2 Patient 2
The patient, 50 years old, presented with a 5-cm mass in the UOQ of the
right breast with N2 disease. Core needle biopsy confirmed the diagnosis
of a triple-negative invasive ductal carcinoma, poorly differentiated with
nodal metastasis. She was interested in breast preservation. Breast cup
size: large B/small C.

Neoadjuvant Therapy
51
(a)
(b)
Figure 3.4.
Mammogram (a) pre-NCT and (b) post-NCT.
(a)
(b)
Figure 3.5.
Four years after completion of radiation therapy.
The patient received NCT (AC-T) with excellent clinical response. The
first images show the mammogram pre- and post-NCT (Figure 3.4).
The patient underwent a lumpectomy and axillary node dissection with
2.5-cm residual disease in the breast. Figure 3.5 shows the patient 4 years
after completion of radiation therapy.

52
Principles and Techniques in Oncoplastic Breast Cancer Surgery
References
1. Fisher B, Anderson S, Bryant J, et al. (2002) Twenty-year follow-up of a
randomized trial comparing total mastectomy, lumpectomy, and lumpec-
tomy plus irradiation for the treatment of invasive breast cancer. N Engl J
Med 347(16):1233–1241.
2. PerloffM, Lesnick GJ. (1982) Chemotherapy before and after mastectomy
in stage III breast cancer. Arch Surg 117(7):879–881.
3. De Lena M, Zucali R, Viganotti G, et al. (1978) Combined chemotherapy-
radiotherapy approach in locally advanced (T3b-T4) breast cancer. Cancer
Chemother Pharmacol 1(1):53–59.
4. Rastogi P, Anderson SJ, Bear HD, et al. (2008) Preoperative chemotherapy:
Updates of National Surgical Adjuvant Breast and Bowel Project Protocols
B-18 and B-27. J Clin Oncol 26(5):778–785.
5. Sachelarie I, Grossbard ML, Chadha M, et al. (2006) Primary systemic ther-
apy of breast cancer. Oncologist 11(6):574–589.
6. Mieog JS, van der Hage JA, van de Velde CJ. (2007) Neoadjuvant chemother-
apy for operable breast cancer. Br J Surg 94(10):1189–1200.
7. Scholl SM, Fourquet A, Asselain B, et al. (1994) Neoadjuvant versus adjuvant
chemotherapy in premenopausal patients with tumours considered too large
for breast conserving surgery: Preliminary results of a randomised trial: S6.
Eur J Cancer 30A(5):645–652.
8. Rouzier R, Mathieu MC, Sideris L, et al. (2004) Breast-conserving surgery
after neoadjuvant anthracycline-based chemotherapy for large breast tumors.
Cancer 101(5):918–925.
9. National Comprehensive Cancer Network Clinical Practice Guidelines:
Breast Cancer. (2011) Version 2.2011.
10. Oh JL, Nguyen G, Whitman GJ, et al. (2007) Placement of radiopaque clips
for tumor localization in patients undergoing neoadjuvant chemotherapy and
breast conservation therapy. Cancer 110(11):2420–2427.
11. Whitman GJ, Strom EA. (2009) Workup and staging of locally advanced
breast cancer. Semin Radiat Oncol 19(4):211–221.
12. Berg WA, Gilbreath PL. (2000) Multicentric and multifocal cancer: Whole-
breast US in preoperative evaluation. Radiology 214(1):59–66.
13. Moon WK, Noh DY, Im JG. (2002) Multifocal, multicentric, and contralat-
eral breast cancers: Bilateral whole-breast US in the preoperative evaluation
of patients. Radiology 224(2):569–576.
14. Krishnamurthy S, Sneige N, Bedi DG, et al. (2002) Role of ultrasound-guided
fine-needle aspiration of indeterminate and suspicious axillary lymph nodes
in the initial staging of breast carcinoma. Cancer 95(5):982–988.
15. Shanta V, Swaminathan R, Rama R, Radhika R. (2008) Retrospective
analysis of locally advanced noninflammatory breast cancer from Chennai,
South India, 1990–1999. Int J Radiat Oncol Biol Phys 70(1):51–58.
16. Londero V, Bazzocchi M, Del Frate C, et al. (2004) Locally advanced breast
cancer: Comparison of mammography, sonography and MR imaging in eval-
uation of residual disease in women receiving neoadjuvant chemotherapy.
Eur Radiol 14(8):1371–1379.

Neoadjuvant Therapy
53
17. Hylton N. (2006) MR imaging for assessment of breast cancer response
to neoadjuvant chemotherapy. Magn Reson Imaging Clin N Am 14(3):
383–389, vii.
18. Le-Petross HC, Hylton N. (2010) Role of breast MR imaging in neoadjuvant
chemotherapy. Magn Reson Imaging Clin N Am 18(2):249–258, viii–ix.
19. Loo CE, Teertstra HJ, Rodenhuis S, et al. (2008) Dynamic contrast-enhanced
MRI for prediction of breast cancer response to neoadjuvant chemotherapy:
Initial results. AJR Am J Roentgenol 191(5):1331–1338.
20. Zakhireh J, Gomez R, Esserman L. (2008) Converting evidence to practice:
A guide for the clinical application of MRI for the screening and management
of breast cancer. Eur J Cancer 44(18):2742–2752.
21. Houssami N, Hayes DF. (2009) Review of preoperative magnetic reso-
nance imaging (MRI) in breast cancer: Should MRI be performed on all
women with newly diagnosed, early stage breast cancer? CA Cancer J Clin
59(5):290–302.
22. Liedtke C, Mazouni C, Hess KR, et al. (2008) Response to neoadjuvant
therapy and long-term survival in patients with triple-negative breast cancer.
J Clin Oncol 26(8):1275–1281.
23. Carey LA, Dees EC, Sawyer L, et al. (2007) The triple negative paradox:
Primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res
13(8):2329–2334.
24. Wang S, Yang H, Tong F, et al. (2009) Response to neoadjuvant therapy
and disease free survival in patients with triple-negative breast cancer. Gan
To Kagaku Ryoho 36(2):255–258.
25. Garber JE, Richardson A, Harris LN, et al. (2006) Neoadjuvant cisplatin in
“triple negative” breast cancer. Breast Cancer Res Treat 100:S149 (Abstract
3074).
26. Gronwald J, Byrski T, Huzarski T, et al. (2009) Neoadjuvant therapy with
cisplatin in BRCA1-positive breast cancer patients. J Clin Oncol 27:15s
(Abstract 502).
27. Buzdar AU, Ibrahim NK, Francis D, et al. (2005) Significantly higher patho-
logic complete remission rate after neoadjuvant therapy with trastuzumab,
paclitaxel, and epirubicin chemotherapy: Results of a randomized trial in
human epidermal growth factor receptor 2-positive operable breast cancer.
J Clin Oncol 23(16):3676–3685.
28. Gianni L, Eiermann W, Semiglazov V, et al. (2010) Neoadjuvant chemother-
apy with trastuzumab followed by adjuvant trastuzumab versus neoadjuvant
chemotherapy alone, in patients with HER2-positive locally advanced breast
cancer (the NOAH trial): A randomised controlled superiority trial with a
parallel HER2-negative cohort. Lancet 375(9712):377–384.
29. Paluch-Shimon S, Wolf I, Goldberg H, et al. (2008) High efficacy of
pre-operative trastuzumab combined with paclitaxel following doxoru-
bicin & cyclophosphamide in operable breast cancer. Acta Oncol 47(8):
1564–1569.
30. Peintinger F, Buzdar AU, Kuerer HM, et al. (2008) Hormone receptor sta-
tus and pathologic response of HER2-positive breast cancer treated with
neoadjuvant chemotherapy and trastuzumab. Ann Oncol 19(12):2020–2025.

54
Principles and Techniques in Oncoplastic Breast Cancer Surgery
31. Straver ME, Rutgers EJ, Russell NS, et al. (2009) Towards rational axillary
treatment in relation to neoadjuvant therapy in breast cancer. Eur J Cancer
45(13):2284–2292.
32. Mamounas EP. (2008) Facilitating breast-conserving surgery and preventing
recurrence: Aromatase inhibitors in the neoadjuvant and adjuvant settings.
Ann Surg Oncol 15(3):691–703.
33. Buzdar AU, Valero V, Theriault R, et al. (2003) Pathological complete
response (pCR) to chemotherapy is related to hormone receptor status. 26th
San Antonio Breast Cancer Symposium (Abstract 302).
34. Tan MC, Al Mushawah F, Gao F, et al. (2009) Predictors of complete patho-
logical response after neoadjuvant systemic therapy for breast cancer. Am J
Surg 198(4):520–525.
35. Ellis MJ, Ma C. (2007) Letrozole in the neoadjuvant setting: The P024 trial.
Breast Cancer Res Treat 105 (Suppl 1) 33–43.
36. Smith IE, Dowsett M, Ebbs SR, et al. (2005) Neoadjuvant treatment of
postmenopausal breast cancer with anastrozole, tamoxifen, or both in com-
bination: The Immediate Preoperative Anastrozole, Tamoxifen, or Com-
bined with Tamoxifen (IMPACT) multicenter double-blind randomized trial.
J Clin Oncol 23(22):5108–5116.
37. Howell A, Cuzick J, Baum M, et al. (2005) Results of the ATAC (Arim-
idex, Tamoxifen, Alone or in Combination) trial after completion of 5 years’
adjuvant treatment for breast cancer. Lancet 365(9453):60–62.
38. Kurosumi M, Takatsuka Y, Watanabe T, et al. (2008) Histopathologi-
cal assessment of anastrozole and tamoxifen as preoperative (neoadjuvant)
treatment in postmenopausal Japanese women with hormone receptor-
positive breast cancer in the PROACT trial. J Cancer Res Clin Oncol
134(6):715–722.
39. Barnadas A, Gil M, Gonzalez S, et al. (2009) Exemestane as primary treat-
ment of oestrogen receptor-positive breast cancer in postmenopausal women:
A phase II trial. Br J Cancer 100(3):442–449.
40. Mlineritsch B, Tausch C, Singer C, et al. (2008) Exemestane as primary
systemic treatment for hormone receptor positive post-menopausal breast
cancer patients: A phase II trial of the Austrian Breast and Colorectal Cancer
Study Group (ABCSG-17). Breast Cancer Res Treat 112(1):203–213.
41. Takei H, Suemasu K, Inoue K, et al. (2008) Multicenter phase II trial
of neoadjuvant exemestane for postmenopausal patients with hormone
receptor-positive, operable breast cancer: Saitama Breast Cancer Clinical
Study Group (SBCCSG-03). Breast Cancer Res Treat 107(1):87–94.
42. Tubiana-Hulin M, Becette V, Bieche I, et al. (2007) Exemestane as neoadju-
vant hormonotherapy for locally advanced breast cancer: Results of a phase II
trial. Anticancer Res 27(4C):2689–2696.
43. Akashi-Tanaka S, Shimizu C, Ando M, et al. (2009) 21-Gene expression pro-
file assay on core needle biopsies predicts responses to neoadjuvant endocrine
therapy in breast cancer patients. Breast 18(3):171–174.

Neoadjuvant Therapy
55
44. Boughey JC, Wagner J, Garrett BJ, et al. (2009) Neoadjuvant chemotherapy
in invasive lobular carcinoma may not improve rates of breast conservation.
Ann Surg Oncol 16(6):1606–1611.
45. Bollet MA, Savignoni A, Pierga JY, et al. (2008) High rates of breast con-
servation for large ductal and lobular invasive carcinomas combining multi-
modality strategies. Br J Cancer 98(4):734–741.
46. Loibl S, von Minckwitz G, Raab G, et al. (2006) Surgical procedures after
neoadjuvant chemotherapy in operable breast cancer: Results of the GEPAR-
DUO trial. Ann Surg Oncol 13(11):1434–1442.
47. Regano S, Hernanz F, Ortega E, et al. (2009) Oncoplastic techniques
extend breast-conserving surgery to patients with neoadjuvant chemother-
apy response unfit for conventional techniques. World J Surg 33(10):
2082–2086.
48. Kirstein LJ, Rafferty E, Specht MC, et al. (2008) Outcomes of multiple wire
localization for larger breast cancers: When can mastectomy be avoided?
J Am Coll Surg 207(3):342–346.
49. Mamounas EP, Brown A, Anderson S, et al. (2005) Sentinel node biopsy
after neoadjuvant chemotherapy in breast cancer: Results from National
Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol
23(12):2694–2702.
50. Kilbride KE, Lee MC, Nees AV, et al. (2008) Axillary staging prior to neoad-
juvant chemotherapy for breast cancer: Predictors of recurrence. Ann Surg
Oncol 15(11):3252–3258.
51. Krag DN, Anderson SJ, Julian TB, et al. (2007) Technical outcomes of
sentinel-lymph-node resection and conventional axillary-lymph-node dissec-
tion in patients with clinically node-negative breast cancer: Results from the
NSABP B-32 randomised phase III trial. Lancet Oncol 8(10):881–888.
52. Xing Y, Foy M, Cox DD, et al. (2006) Meta-analysis of sentinel lymph node
biopsy after preoperative chemotherapy in patients with breast cancer. Br J
Surg 93(5):539–546.
53. Sakakibara M, Nagashima T, Kadowaki M, et al. (2009) Clinical significance
of axillary microresiduals after neoadjuvant chemotherapy in breast can-
cer patients with cytologically proven metastases. Ann Surg Oncol 16(9):
2470–2478.
54. Hunt KK, Yi M, Mittendorf EA, et al. (2009) Sentinel lymph node surgery
after neoadjuvant chemotherapy is accurate and reduces the need for axillary
dissection in breast cancer patients. Ann Surg.
55. Chen AM, Meric-Bernstam F, Hunt KK, et al. (2005) Breast conservation
after neoadjuvant chemotherapy. Cancer 103(4):689–695.
56. Akay C, Meric-Bernstam F, Grubbs LG, et al. (2011) Validation of prog-
nostic index score for local-regional recurrence after mastectomy or breast
conservationtherapy in patients treated with NCT. 64th Annual Cancer
Symposium, Society Of Surgical Oncology, San Antonio TX (March 2011)
(Abstract 34).

56
Principles and Techniques in Oncoplastic Breast Cancer Surgery
57. Fisher B, Bryant J, Wolmark N, et al. (1998) Effect of preoperative
chemotherapy on the outcome of women with operable breast cancer. J Clin
Oncol 16(8):2672–2685.
58. Kuerer HM, Newman LA, Smith TL, et al. (1999) Clinical course of breast
cancer patients with complete pathologic primary tumor and axillary lymph
node response to doxorubicin-based neoadjuvant chemotherapy. J Clin Oncol
17(2):460–469.
59. O’Shaughnessy JA, Osborne CK, Pippen JE, Jr., et al. (2009) Efficacy of
BSI-201, a poly (ADP-ribose) polymerase-1 (PARP1) inhibitor, in combina-
tion with gemcitabine/carboplatin (G/C) in patients with metastatic triple-
negative breast cancer (TNBC): Results of a randomized phase II trial. J Clin
Oncol 27:18s (Abstract 3).
60. Fong PC, Boss DS, Carden CP, et al. (2008) AZD2281 (KU-0059436),
a PARP (poly ADP-ribose polymerase) inhibitor with single agent anti-
cancer activity in patients with BRCA deficient ovarian cancer: Results from
a phase I study. J Clin Oncol 26 (Abstract 5510).
61. Yap TA, Boss DS, Fong PC, et al. (2007) First in human phase I phar-
macokinetic (PK) and pharmacodynamic (PD) study of KU-0059436 (Ku),
a small molecule inhibitor of poly ADP-ribose polymerase (PARP) in cancer
patients (p), including BRCA1/2 mutation carriers. J Clin Oncol 25(18S):
3529.
62. IsakoffSJ. (2010) Triple-negative breast cancer: Role of specific chemother-
apy agents. Cancer J 16(1):53–61.
63. Mauriac L, MacGrogan G, Avril A, et al. (1999) Neoadjuvant chemotherapy
for operable breast carcinoma larger than 3 cm: A unicentre randomized trial
with a 124-month median follow-up. Institut Bergonie Bordeaux Groupe Sein
(IBBGS). Ann Oncol 10(1):47–52.
64. Rouzier R, Extra JM, Carton M, et al. (2001) Primary chemotherapy for
operable breast cancer: Incidence and prognostic significance of ipsilat-
eral breast tumor recurrence after breast-conserving surgery. J Clin Oncol
19(18):3828–3835.
65. Bonadonna G, Valagussa P, Brambilla C, et al. (1998) Primary chemother-
apy in operable breast cancer: Eight-year experience at the Milan Cancer
Institute. J Clin Oncol 16(1):93–100.
66. Chen AM, Meric-Bernstam F, Hunt KK, et al. (2004) Breast conservation
after neoadjuvant chemotherapy: The MD Anderson Cancer Center. J Clin
Oncol 22(12):2303–2312.
67. Cance WG, Carey LA, Calvo BF, et al. (2002) Long-term outcome of
neoadjuvant therapy for locally advanced breast carcinoma: Effective clin-
ical downstaging allows breast preservation and predicts outstanding local
control and survival. Ann Surg 236(3):295–302 (Discussion 3).
68. Beriwal S, Schwartz GF, Komarnicky L, Garcia-Young JA. (2006) Breast-
conserving therapy after neoadjuvant chemotherapy: Long-term results.
Breast J 12(2):159–164.

Chapter 4
SKIN INCISIONS
Mahmoud El-Tamer
Breast Service, Department of Surgery,
Memorial Sloan-Kettering Cancer Center,
New York, NY 10065, USA
As of today, surgical resection remains the mainstay of breast cancer
management. Multiple percutaneous and transcutaneous therapeutic
modalities are currently being explored, but all remain experimental at
this time. In order to achieve a surgical resection of a breast tumor, skin
incisions remain mandatory. The long-term appearance of such incisions is
crucial to the final cosmetic result. When planning an incision, the surgeon
must keep multiple considerations in mind:
(1) Convenience for appropriate tumor removal
(2) Access to repair the defect
(3) Good cosmetic placement
(4) Allowance of further surgery (mastectomy without major inconve-
nience)
In the upper quadrants, we use curvilinear skin incisions most frequently.
Curvilinear incisions placed in the lines of Kraissl1 result in optimal heal-
ing and minimal scar separation. Furthermore, curvilinear incisions in the
upper quadrant are optimal when resection of the tumor necessitates the
removal of the overlying skin. After repair of an elliptical incision in
the upper quadrant, the nipple areola is lifted, thereby enhancing the cos-
metic result in a ptotic breast. Every effort should be made while designing
the elliptical skin incision to avoid any displacement of the nipple, either
laterally or medially. Misplacing the nipple areola complex may be more
common in lesions that are on either side of the 12 o′clock axis.
57

58
Principles and Techniques in Oncoplastic Breast Cancer Surgery
Figure 4.1.
Guidance for appropriate nipple position.
We recommend that surgeons review the nipple meridian (Chapter 2)
for guidance regarding the appropriate position of the nipple (Figure 4.1).
Excessive skin resection in the upper quadrant may result in the upward
displacement of the nipple areola. We also try to avoid extending upper
quadrant curvilinear incisions beyond the 3 o′clock and 9 o′clock axes. Such
extensions may result in a suboptimal cosmetic result as scars traversing
the 3 o′clock or 9 o′clock axes may widen or retract the NAC medially or
laterally.
Radial incisions in the upper outer quadrant that are at the 10 o′clock
axis on the right and the 2 o′clock axis on the left allow large resections
of skin and underlying breast tissue with excellent repair. Such radial inci-
sions may extend from the areolar edge up to the axilla, facilitating the
performance of a sentinel node biopsy or even a complete axillary node
dissection. Closure of these incisions does not usually result in the lateral
retraction of the nipple areola complex. We limit radial incision in the upper
outer quadrant to tumors that would otherwise require a quadrantectomy
(Figure 4.2).2
Radial incisions can be performed in the upper inner quadrants of the
breast. We believe, however, that they result in unsightly scars that are
difficult to hide with a brassiere. We avoid placing any radial incisions
between the 11 o′clock and 3 o′clock axes on the right side, and the 10 o′clock
and 1 o′clock axes on the left (Figure 4.3).

Skin Incisions
59
(a)
(b)
Figure 4.2.
Radial incision in the upper outer quadrant limited to tumors that would
otherwise require a quadrantectomy.
Figure 4.3.
We avoid placing any radial incisions in the areas delineated in this drawing
as they may result in unsightly scars.
In the lower quadrants, curvilinear incisions are applicable, particularly
with small tumors. When the skin overlying a tumor is being excised as
part of the resection, curvilinear incision should be avoided in the lower
quadrants. Approximation of the skin after resection of a curvilinear ellipti-
cal skin incision in the lower quadrant will result in inferior displacement of
the nipple areola and a corresponding unsatisfactory cosmetic appearance,
better known as “Bird’s Beak”.3 To avoid this type of cosmetic deformity,
we use radial elliptical incision in the lower quadrants exclusively when

60
Principles and Techniques in Oncoplastic Breast Cancer Surgery
(a)
(b)
(c)
Figure 4.4.
Radial incisions. (a) Curvilinear incision. (b) Radial incision. (c) Incision
with skin excision.
planning skin resection. We also prefer radial incisions when faced with
large-volume resection in the lower quadrants, as they facilitate repair of
the defect in a mastopexy fashion, enhancing the final cosmetic outcome
(Figure 4.4).
4.1 Periareolar Incision
In our experience, skin incisions performed around the areola provide the
best cosmetic results. These wounds heal with minimal scarring, and due to

Skin Incisions
61
Figure 4.5.
Periareolar incision.
their location, fade well, particularly when placed exactly at the edge of the
areola. We recommend drawing the incision prior to prepping the patient
because the edge of the areola may become blurry following painting with
antiseptic solutions. This is particularly the case when using iodine-based
solutions and in patients with fair skin tone coloring.
Areola size must be taken into consideration when planning periareolar
incisions (Figure 4.5). Access to resection and repair may be limited in
patients with areolar diameters of 2 cm or less. In this particular group,
we recommend extending the incision laterally, medially, or both, at the
3 o′clock or 9 o′clock axes. Such extensions enhance access to and allow
appropriate repair of the defect.
Maintaining the blood supply of the nipple areola complex during sur-
gical intervention is crucial. We have rarely had any significant tissue loss
with periareolar skin incisions, irrespective of the location of the incision,
provided we have limited the skin incision to less than 50% of the circumfer-
ence of the areola. During the surgical procedure, aggressive and prolonged
retraction of the areola may result in significant damage to the blood sup-
ply, particularly in heavy smokers. We frequently release the traction on
the areola to allow the flow of blood through the nipple areola complex.
We also recommend judicious use of the electrocautery to the base of the
nipple areola, particularly when the nipple areola complex is raised as a
dermal flap. Patients frequently complain of loss of some of sensation with
circumareolar incisions in the immediate perioperative period; however,

62
Principles and Techniques in Oncoplastic Breast Cancer Surgery
the vast majority of patients report normal sensations 3 to 6 months after
surgery.
4.2 Inframammary Incision
Incisions along the inframammary line have excellent cosmetic results, par-
ticularly in ptotic breasts. The scars are very well hidden, and patients
frequently can only visualize the scars in a mirror. The scars allow good
access to the breast and the possibility of utilizing the subcutaneous tissue
below the inframammary line for reconstruction. One of the major incon-
veniences that arise with such incisions is their relative inconvenience when
planning a completion mastectomy.
4.3 Skin Closure
We have made it a routine to approximate the skin in 2 layers, using
absorbable suture. The first layer of closure is performed by approximat-
ing the deep dermis with interrupted and inverted 4-0 filaments. The
sutures are applied at 1 cm to 2 cm intervals. We approximate the skin
with running subcuticular 4-0 monofilament absorbable sutures. Dressing
is accomplished with steri-strips with an overlying transparent imperme-
able adhesive. We always buttonhole the impermeable dressing to allow
extrusion of the nipple (Figure 4.6).
Figure 4.6.
A buttonholed impermeable dressing allowing extrusion of the nipple.

Skin Incisions
63
References
1. Kraissl CJ. (1951) The selection of appropriate lines for elective surgical inci-
sions. Plast Reconstr Surg (1946) 8(1):1–28.
2. Veronesi U, Zurrida S. (2000) Quadrantectomy for malignant disease. Operat
Tech Gen Surg 2:132–136.
3. Clough KB, Kaufman GJ, Nos C, et al. (2010) Improving breast cancer
surgery: A classification and quadrant per quadrant atlas for oncoplastic
surgery. Ann Surg Oncol 17(5):1375–1391.

This page intentionally left blank
This page intentionally left blank

Chapter 5
MARGINS IN BREAST CANCER SURGERY
Adriana D. Corben∗and Monica Morrow†
∗Department of Pathology,
Memorial Sloan-Kettering Cancer Center,
New York, NY10065, USA
†Breast Service, Department of Surgery,
Memorial Sloan-Kettering Cancer Center,
New York, NY10065, USA
Breast-conserving therapy (BCT), consisting of the removal of the pri-
mary tumor and whole breast irradiation, is a well-accepted method for
breast cancer management. Six prospective randomized trials, some with
follow-up of 20 years or more,1,2 have established without a doubt that
there is no survival advantage for mastectomy compared with the more
limited surgery and radiation therapy (RT). Over time, the rates of local
recurrence after BCT have declined steadily and are now considerably less
than 10% at 10 years of follow-up.3,4 However, the appropriate extent of
surgical resection needed to maintain local control remains a matter of
debate, and the lack of consensus regarding what constitutes an adequate
negative margin results in multiple trips to the operating room for margin
re-excision in a significant number of patients, and unnecessary mastec-
tomies in others.5 The perception that a more widely clear margin enhances
local control has been a major factor in the development of oncoplastic
surgery.
The demonstration in the Early Breast Cancer Trialists Collaborative
Group (EBCTCG) overview that differences in local control between treat-
ments of 10%–20% at 5 years are associated with significant differences in
breast cancer-specific survival at 15 years6 has focused new attention on the
importance of local control. For many years, disease burden as defined by
65

66
Principles and Techniques in Oncoplastic Breast Cancer Surgery
margin status was felt to be the primary determinant of local control. Over
time, it has become increasingly clear that both the underlying genetics of
the tumor and the availability of effective systemic therapy are also critical
components of local control. In this chapter, we will discuss techniques of
margin evaluation, review the available data on the relationship between
margin status and local control for invasive and intraductal cancer, and dis-
cuss the impact of molecular subtype of breast cancer and targeted therapy
on local control outcomes.
5.1 Methods of Margin Evaluation
The term “margin” is often used without specifying the pathologic method
used to determine the margin status. Unfortunately, there is neither a stan-
dard method of margin evaluation nor a standard number of histologic
sections that are examined from each margin surface. Margins can be eval-
uated using a radial method, shaved method, or by shaving the walls of
the lumpectomy cavity. These techniques are described later. Variations
in pathology protocols take on considerable potential significance when
clinicians assert that differences of 1–2 mm separate an adequate from an
inadequate resection. This variability in margin processing is documented
in a study which examined 91 consecutive excisional breast biopsies submit-
ted from 50 different hospitals to a single university pathology department
over a 2-year period.7 Only 18% of reports described the technique (shaved
or radial) used to evaluate the margins. Only 30% of the specimens were
submitted in total; an unknown amount of tissue was submitted in 1%, and
representative sections averaging 13 blocks per specimen were submitted
in 69%.
5.1.1 Radial Margin Approach
The most common method of margin assessment is the perpendicular (or
radial) margin technique, which allows for precise measurement of the dis-
tance separating the tumor from the inked margin. With this method, the
specimen is received with at least 2 of the margins marked with clips or
sutures to allow specimen orientation. The 6 margins of the specimen are
inked in 6 different colors by the prosector. The ink should be applied gen-
tly so that no artificial disruption occurs. Excess ink should be carefully
removed by pressing with gauze to reduce the problem of “running” ink.
A brief application of acetone on the inked surface helps stain enhancement
and reduces the problem of running ink. The inked specimen is sequen-
tially cut into 0.2–0.3 cm slices, perpendicular to its long axis, so that the

Margins in Breast Cancer Surgery
67
A. Radial
B. En face
Black: 0.8 cm; Yellow: 0.3 cm
Blue: 0.6 cm
Superior
Margin
Superior
Margin
Submit sections to document all margins measuring
1.0 cm or less
Figure 5.1.
(a) Radial (perpendicular) margin evaluation. With the radial margin tech-
nique, each margin surface is inked a different color, and the distance from the tumor
to the ink is reported for each margin. (b) Shaved (en face) margin evaluation. With
the en face method, the entire surface is shaved and no inking is needed.
perimeter of each tissue slice contains a few (2 to 4) margins identified by
the different ink colors. Margins within 0.5 cm from the index lesion are
best evaluated in their entirety, and the more distant margins can be repre-
sentatively sampled [Figure 5.1(a)]. With this method, the pathologist can
report the exact microscopic distance from the tumor to each individual
margin and distinguish between a truly positive margin (tumor at ink) and
a close margin. The extent of tumor at the margin or in close proximity to
it can also be assessed.
The disadvantages of this method include running ink, imprecise margin
orientation, and surface complexity. Running of the ink from the irregular
fatty specimen surface to the inside of the specimen [Figure 5.2(a)], and dif-
ferent color inks running into each other [Figure 5.2(b)], occur frequently,
leading to possible overinterpretation and false-positive margins. In addi-
tion, the breast tissue is soft and may be artificially disrupted when the
specimen is compressed to obtain a specimen radiograph or during gross
examination. The flattening of the specimen in the gross room may distort
the margin orientation. In one study, surgeons and pathologists disagreed
on margin designation in 31% of cases when 2 marking sutures for orienta-
tion were placed in the operating room. The presence of skin or muscle on
the specimen did not reduce the disorientation rate, but the specimen size
was highly correlated with this problem. Specimens less than 20 cm in size

68
Principles and Techniques in Oncoplastic Breast Cancer Surgery
(a)
(b)
Figure 5.2.
Problems with radial margin assessment and inking. (a) “Running” of
ink (blue arrow) inside a specimen challenges the interpretation of the distance of the
ductal carcinoma in situ (black arrows) from the true inked resection margin (red arrow).
(b) Different color inks (orange, blue and green arrows) running into each other create
false interpretations of margin designation.
had a disorientation rate of 78% compared to only 20% for larger speci-
mens (p < 0.001).8 In addition, it has been estimated that the complete
examination of a 2 cm spherical specimen would require in excess of 3,000
sections.9 Taking into account that the surfaces of breast excisions are
highly complex and irregular, and that the surface area of some specimens
is quite large, margin assessment, using the inked radial margin method,
is highly subject to sampling error.
5.1.2 Shaved Margin Approach
The use of shaved (en face) margins allows the oriented specimen to be
inked entirely in one color as long as the prosector is able to maintain
the proper orientation. This eliminates the problem of different colored
inks running together. The margins are shaved offparallel to the outer
surface of the specimen, similar to the process of peeling an orange, at a
tissue depth of 0.2–0.3cm [Figure 5.1(b)]. The shaved margins closest to
the index lesion are submitted entirely. The sections are embedded en face
with the inked surface facing downward so that the microscopic examina-
tion starts from the inner aspect of the tissue. With this method, a margin
is reported as positive when the tumor is present anywhere in the sec-
tion. This means that malignant cells may be present within a 0.2–0.3 cm
radius from the surgical margin or at the margin itself, but that the exact
distance of the tumor to margin cannot be assessed. If no tumor is iden-
tified, the margin is reported as negative. The advantages of this method

Margins in Breast Cancer Surgery
69
include a straightforward microscopic examination, no occurrence of ink
problems, and the examination of a large proportion of the specimen’s sur-
face with relatively few sections. Although advantageous to the pathologist,
this technique is extremely problematic for surgeons because it substan-
tially increases the number of margins which are called positive.
Guidi et al.10 compared the shaved margin technique with the con-
ventional radial margin method by evaluating 22 surgical specimens using
both methods. The specimens were inked, and the margins were shaved
and microscopically examined. The tissue was then extracted from the
block, cut perpendicular to the inked surface, and re-embedded to assess
the radial inked margin status. The study demonstrated that a negative
shaved margin was highly predictive of a negative inked margin (98% con-
cordance), but the positive predictive value was much lower, with only 61%
of the positive shaved margins being called positive (tumor at ink) by the
radial margin technique. In a study at Memorial Sloan-Kettering Cancer
Center (MSKCC), Wright et al. reported that when the MSKCC Depart-
ment of Pathology switched from radial margin assessment to the shaved
method, the positive margin rate increased from 16% to 49% even though
the surgical technique did not change.11
5.1.3 Cavity Shave Approach
The evaluation of separate cavity shaves, obtained by the surgeon, pro-
vides a solution to margin assessment which combines the advantages of
both the radial and the shaved margins. With this approach (Figure 5.3),
the surgeon resects the index lesion and then takes separate shaved mar-
gins from the surgical cavity. A separate anterior margin may or may
not be submitted. The main specimen, containing the tumor, is received
unoriented and does not necessarily need to be inked. The tumor is
entirely sampled, including any prior biopsy site, and 1 or 2 represen-
tative sections of any grossly uninvolved breast tissue present are also
submitted. Each shave specimen represents a margin (medial, lateral, infe-
rior, superior, and posterior walls of the surgical cavity) and is received
oriented with a suture or metal clip, designating the final margin sur-
face. Sometimes, one specimen may consist of 2 adjacent margins. Each
margin is inked on the side designated by the suture or clip, perpen-
dicularly sectioned, and submitted either entirely or representatively in
10 blocks (Figure 5.3). In our institution, after an internal review, only
6 blocks are now submitted per shave specimen unless any epithelial
atypia or carcinoma is microscopically identified; the remaining margin
is then entirely submitted. This technique allows precise margin desig-
nation and accurate measurement of the margin width. These specimens

70
Principles and Techniques in Oncoplastic Breast Cancer Surgery
Radial sections
“Full face” section of the tumor
Shaved margins
from cavity site
Lumpectomy specimen
Figure 5.3.
Cavity shaved margin evaluation (1 out of 5 shown). With the cavity margin
evaluation method, the primary lumpectomy specimen is not inked. The outer surface
of each of the individual margins is inked, and the distance to any tumor measured.
are more easily handled by the prosector with limited manipulation of
the tissue, and the use of the same color ink contributes to a reduction
of problematic artifacts of the prior described techniques. A significant
increase in the number of blocks and slides is the main disadvantage
of this method for most institutions. A number of investigators have
reported that this method significantly reduces the rate of re-excision for
close margins.12−18 These results are summarized in Table 5.1. One issue
with the routine use of cavity shave margins is the potential removal of
larger amounts of breast tissue, resulting in a worsening of cosmetic out-
come. Huston et al. compared the rates of margin positivity and volume
of tissue, excised among patients undergoing lumpectomy, with shaving
of all margins (n = 45), lumpectomy with selective shaving of margins
thought by the surgeon to be suspicious (n = 77), and lumpectomy alone
(n = 49). The routine resection of additional margins resulted in the
lowest re-operation rate, 17.7%, compared to 32.5% and 38.7% for selec-
tive re-excision and lumpectomy, respectively. The mean total specimen
volumes were 129 cm,3 46 cm,3 and 37 cm3 for the 3 groups.12 In contrast,

Margins in Breast Cancer Surgery
71
Table 5.1.
Impact of tumor cavity sampling on re-excision rates.
% Positive margin
Author
# Cases
Cavity margins
Routine
P -value
Huston TL12
171
18
39
Jacobson AF13
125
18
66
Tengher-Barna I17
107
13
33
Marudanayagam R15
786
5.6
12.5
<0.01
Rizzo M16
320
15
43
<0.05
Kobberman A14
138
22
42
0.01
Rizzo et al.16 found no difference in the volume removed with or without
additional shaved margins if negative margins were achieved with the initial
excision. However, in patients requiring re-excision, a smaller total volume
of breast tissue was removed if the cavity shave technique was used initially
(p < 0.005).
5.1.4 Other Issues Related to Margin Assessment
In addition to the method of margin assessment, other factors related to
specimen processing may influence the margin width or whether or not
the tumor is identified at the margin. Graham et al. compared the mea-
surement of the mean height of the lumpectomy specimen in the oper-
ating room by the surgeon to the measurement in the pathology lab in
100 consecutive specimens19 and found a 46% decrease from a mean of
2.6 cm as measured by the surgeon to 1.4 cm as measured in pathology.
This was independent of patient age, lesion type (mass versus calcifica-
tions) or breast density. When compression devices were used for specimen
radiography, the decrease was 54% compared to 41% when these devices
were not used (p = 0.0003). These findings indicate that measurements of
anterior and posterior margin distances on the primary tumor specimen
are subject to significant variation depending upon how the specimen is
handled.
5.2 Margin Width and Local Recurrence: Invasive Cancer
Considering the lack of standardization in pathology methods, it is not
surprising that there is little consensus regarding what constitutes an
adequate negative margin. Azu et al.20 surveyed a population-based sample

72
Principles and Techniques in Oncoplastic Breast Cancer Surgery
of 318 surgeons identified from patients in the Surveillance, Epidemiol-
ogy, and End Results (SEER) registry. Surgeons were asked “What neg-
ative margin width precludes the need for re-excision in a 60-year-old
with a 0.8 cm invasive cancer which is estrogen receptor (ER), proges-
terone receptor (PR), and HER2 negative?” The following four options
were offered: Tumor not touching ink; >1–2 mm; >5 mm; or >10 mm.
No answer was endorsed by more than 50% of respondents. Only 11%
selected the tumor not touching ink option, 42% selected >1–2mm, and
19% selected >10 mm. Similar variation exists among radiation oncolo-
gists. In a survey of 1,133 North American and European radiation oncol-
ogists, 45% of those from North America endorsed a margin of tumor
not touching ink, while those in Europe favored more widely clear mar-
gins, with greater than 5 mm being the most common answer, selected
by 29%.21 The lack of consensus on what constitutes an adequate neg-
ative margin results in the frequent performance of re-excision to obtain
more widely clear margins. Morrow et al.,5 reporting on a population-based
sample from the SEER registry of 800 women attempting BCT, observed
that although the procedure was successful in 88%, 22% underwent a re-
excision. Other studies report a wide variation in re-excision rates ranging
from 6%15 to 49%,22 with the majority noting re-excision in 15% to 30%
of patients.12,23,24
The prospective randomized trials that established the safety and effi-
cacy of BCT1,2,25−28 do not provide much guidance on the margin question
because only the National Surgical Adjuvant Breast and Bowel Project
(NSABP) B06 trial used a microscopic definition of a negative margin,
which was tumor not touching ink.1 Although the other trials are often
perceived as requiring more widely clear margins, they relied upon gross
margin definitions, making the actual margin width impossible to assess.
Similarly, although a trial by Veronesi et al., which randomized patients
to quadrantectomy or a more limited tumorectomy, demonstrated a lower
rate of local recurrence in the quadrantectomy group (2.2% versus 7.0%),
this study also relied on gross margin assessment. The tumorectomy was
performed with a gross margin of 1 cm, but in a subset of patients who had
microscopic margin evaluation, 16% of those in the tumorectomy group
had positive margins.29 The uncertainty over margin status in this trial
makes it impossible to conclude that a larger quadrantectomy-type proce-
dure is associated with a lower rate of local recurrence than a more limited
resection with negative inked margins. The authors have been unable to
identify any prospective randomized trials that directly compare outcomes
between microscopically verified margins of differing widths.
Recently, Houssami et al. performed a methodologically rigorous meta-
analysis of the relationship between local recurrence and margin distance.30

Margins in Breast Cancer Surgery
73
The analysis included 21 studies with 14,571 patients and 1,026 local
recurrences. The relationship between margin status and local recurrences
was verified, with an odds ratio (OR) for local recurrence of 2.42 for positive
versus negative margins (p < 0.001). No relationship between margin
width, defined as 1 mm versus 2 mm versus 5 mm, and local recurrence
was identified. Although a trend for a benefit of more widely clear margins
was seen in some models, this was not statistically significant after adjust-
ing for other treatment factors, such as the use of a radiation boost or
receipt of endocrine therapy. This meta-analysis, which included informa-
tion on a large number of factors relevant to local recurrences, such as date
of study enrollment, patient age, use of radiation including a boost, and
pathologic tumor features such as lymphovascular invasion (LVI), exten-
sive intraductal component (EIC), and tumor grade, provides the most
convincing evidence to date that margins more widely clear than tumor
not touching ink do not have a major impact upon local control in the era
of modern multidisciplinary therapy.
Although this may seem counterintuitive, it becomes much more logical
if one considers that even mastectomy does not eliminate the problem of
local recurrence. In the initial randomized trials comparing BCT to mas-
tectomy in which at least grossly negative margins were required, only
the Milan study,2 which included T1 cancers treated with radical mastec-
tomy, showed a statistically significant reduction in local recurrence for
mastectomy compared to BCT (Table 5.2).1,25,26 This, coupled with the
observation from the EBCTCG overview6 that even with the addition of
postmastectomy RT, the incidence of local recurrence is higher in node-
positive women than it is in node-negative women, indicates that local
recurrence may be due to either excessive tumor burden or aggressive biol-
ogy. The failure to observe a decrease in local recurrence with surgical
margins more widely clear than tumor on ink suggests that once the dis-
ease burden is reduced to this level (i.e., no clinically detectable cancer),
tumor biology is the main determinant of local control.
Table 5.2.
Local recurrence in randomized trials of mastectomy versus BCT.
% Local recurrence
Trial
Follow-up (yrs)
BCT
Mastectomy
Institute Gustave-Roussy25
15
9
14
Milan 12
20
9
2*
NSABP B061
20
14
10
Danish26
6
3
4
∗p < 0.0001. BCT, breast-conserving therapy.

74
Principles and Techniques in Oncoplastic Breast Cancer Surgery
5.2.1 The Influence of Histology on Margin Width
Variations in the growth patterns of different histologic types of cancers
raise the possibility that the same margin width may not be appropriate
for all histologic tumor types. Infiltrating lobular cancers are frequently
multifocal and grow as single cells in linear strands separated by normal
stroma,31 raising the possibility that margins negative only by “tumor not
touching ink” might be associated with a significant residual tumor burden.
However, clinical studies do not document a higher rate of local recurrence
after BCT for lobular cancers compared to ductal cancers,32−34 suggesting
that if negative margins are obtained, the growth pattern is irrelevant.
Galimberti et al.35 analyzed 382 patients with pure infiltrating lobular
carcinoma, treated with BCT, to determine if rates of local control differed
among those with margins less than 1 cm compared to 1 cm or greater. The
local failure rate was 4.6% for the less than 1 cm margin group compared
to 3.7% in the 1 cm or greater group, leading the authors to conclude
that more widely negative margins were not necessary for patients with
infiltrating lobular carcinoma.
The other group of concern is patients with an EIC in association with
their invasive cancer. Early studies, performed prior to the routine inking
of margins, suggested that an EIC was associated with higher rate of local
recurrence in patients undergoing BCT.36 Holland et al. documented that
approximately 30% of patients with EIC-positive cancers had prominent
intraductal carcinoma more than 2 cm beyond the primary tumor, com-
pared to only 2% of patients with EIC-negative tumors,37 indicating that
a substantial number of patients with an EIC, treated by excision to grossly
negative margins, have a heavy residual tumor burden. When patients with
an EIC are excised to negative-inked margins, the rates of local recurrence
are not increased compared with patients without an EIC.38,39 However,
Faverly et al. have demonstrated that low- and intermediate-grade ductal
carcinoma in situ (DCIS) often grows with gaps between the DCIS lesions,
although these gaps are usually less than 5 mm in size,40 suggesting that
margins negative by only tumor not touching ink could be associated with
a significant residual tumor burden. The issue of margins and DCIS is dis-
cussed in detail later in this chapter. The presence of an EIC in association
with invasive cancer indicates it may be prudent to obtain a margin of at
least 2 mm if large amounts of DCIS are in proximity to the margin. In the
case of both infiltrating lobular carcinoma and an EIC, clinical judgment
remains important. A single duct of DCIS or microscopic focus of lobular
carcinoma in close proximity to the margin is unlikely to be associated with
a heavy residual tumor burden, whereas a large area of tumor immediately
adjacent to a margin is associated with a greater risk of residual disease41
and should prompt re-excision.

Margins in Breast Cancer Surgery
75
5.2.2 Other Factors Influencing Local Control in Invasive
Cancer
It is important to recognize that a “negative” margin does not indicate that
there is no residual tumor in the breast. Holland et al.,42 in a landmark
study using serial subgross sectioning to evaluate the remaining breast
tissue in 264 mastectomy specimens from patients with clinically unifocal
cancers 4 cm or less in size, showed that only 39% of cases had no additional
tumor beyond the index cancer. In 20% of cases the tumor foci were within
2 cm of the index tumor, and in 41% of cases the tumor foci were more
than 2 cm from the primary tumor (Figure 5.4). From a practical point
of view, a negative margin indicates that the residual tumor burden in
the breast is low enough that it is likely to be controlled by RT. The role
of RT in maintaining local control is well documented in the EBCTCG
overview.6 At 5 years, the absolute incidence of local recurrence in node-
negative women receiving RT was 16% lower than in those not receiving
RT, while a 30% reduction in isolated local recurrence was seen for node-
positive women. These reductions in local recurrence at 5 years translate
to 15-year survival gains of 5% and 7% in node-negative and node-positive
women, respectively. Bartelink et al., in a prospective randomized trial
examining the benefits of boost dose of RT on local control, demonstrated
statistically significant reductions in local recurrence with the addition of
a boost in women of all ages.43 While the role of RT in local control has
long been recognized, the impact on local control of both improvements in
systemic therapy and its widespread use are less well recognized.
20%
41%
4 cm
2 cm
Multifocality of "Localized" Breast Cancer
39%
Figure 5.4.
Distribution of cancer identified on serial subgross sectioning of 264 mas-
tectomy specimens with clinically localized tumors less than 4 cm in size. Only 39% of
clinically localized cancers had all malignant cells confined to the tumor mass. In 20%
of the cases, the residual cancer was within 2 cm of the primary; in the remainder, it
was within 4 cm. (Data from Holland R et al.42)

76
Principles and Techniques in Oncoplastic Breast Cancer Surgery
The majority of women with invasive breast cancer now receive some
form of adjuvant systemic therapy in addition to surgery and RT. Both
endocrine therapy and chemotherapy significantly reduce the likelihood
of local recurrence after BCT. In the NSABP B14 trial, in which node-
negative, ER positive women were randomized to tamoxifen citrate or
placebo, the 10-year rate of in-breast recurrence was reduced from 14.7%
in the placebo group to 4.3% in the tamoxifen group.44 In the NSABP B13
trial, node-negative ER negative women were randomized to chemotherapy
or a no-treatment control group.45 At 8 years, local recurrence was seen in
only 2.6% of those receiving chemotherapy compared with 13.4% of con-
trols. In a report of 3,799 node-negative women participating in 5 NSABP
trials of adjuvant systemic therapy, the cumulative incidence of in-breast
recurrence at 12 years for those receiving adjuvant therapies was only
6.6%.3 Since the time that these trials were conducted, our ability to target
therapy has improved, and this will undoubtedly result in a further decrease
in local recurrence rates. For example, in the randomized trials that estab-
lished the efficacy of adjuvant trastuzumab, the addition of trastuzumab to
chemotherapy resulted in a 40% decrease in locoregional recurrence com-
pared to treatment with chemotherapy alone.46 Similar results have been
reported in ER positive, node-negative patients when systemic treatment is
selected on the basis of the Oncotype DxTM score. Although this score was
developed to predict the risk of systemic recurrence, Mamounas et al.47
demonstrated that in the absence of systemic treatment, patients with
high-risk Oncotype DxTM scores had an 18.4% risk of locoregional recur-
rence (LRR) compared to those with low-risk scores who had a 10.8% risk.
The addition of tamoxifen, appropriate treatment for those with low-risk
scores, reduced the incidence of LRR by more than 50% to 4.3% in the low-
risk group. In contrast, a much more modest reduction in LRR from 18.4%
to 15.8% was seen in the high-risk group. However, when chemotherapy
was added, the LRR rate in the high-risk group decreased to 7.8%.
The importance of biology and targeted therapy is further supported
by the emerging literature on the impact of tumor subtype on local recur-
rence after BCT or mastectomy. Both Millar et al.48 and Nguyen et al.49
have demonstrated that the rate of local recurrence after BCT varies
among the intrinsic subtypes of breast cancer as approximated by ER,
PR, and HER2 status. In both studies, the lowest rates of local recur-
rence at 5 years were seen among the ER, PR positive, HER2-negative
(Luminal A-like) group, and the highest rates among the triple negative
(basal-like) and ER negative, HER2-positive patients in the absence of
adjuvant trastuzumab. However, ER, PR, and HER2 status are not indica-
tors of the need for more widely clear margins because the same pattern of
an increased risk of chest wall recurrence among ER negative patients,

Margins in Breast Cancer Surgery
77
regardless of HER2 status, was observed in a retrospective analysis of
the Danish Breast Cancer Group randomized trials of mastectomy with
or without RT.50 As previously noted, the addition of trastuzumab to
chemotherapy has been shown to reduce the risk of local recurrence in
HER2-overexpressing patients,46 indicating that targeted therapy is a
major contributor to local control. In aggregate, this information indi-
cates that factors other than disease burden are key determinants of local
control.
5.2.3 Margins After Neoadjuvant Chemotherapy
Prospective randomized trials have demonstrated that the use of preop-
erative (neoadjuvant) chemotherapy allows BCT in patients who would
have required a mastectomy if surgery was performed initially.51,52 A meta-
analysis of randomized trials of neoadjuvant versus adjuvant chemotherapy
demonstrated no increase in the risk of LRR after neoadjuvant therapy as
long as surgery was performed (hazard ratio, 1.12; 95% confidence inter-
val (CI), 0.92–1.37; p = 0.25).53 However, in the NSABP B18 trial, after
9 years of follow-up, the patients who required chemotherapy to down-
stage to lumpectomy (n = 65) had a 15.9% rate of in-breast recurrence
compared to 9.9% for those who were lumpectomy candidates at presenta-
tion (p = 0.04).54 This finding raises the possibility that margins need to
be considered somewhat differently in the setting of neoadjuvant therapy.
The purpose of neoadjuvant therapy is to allow the removal of a smaller
amount of breast tissue than would be necessary if initial surgery was per-
formed. Breast cancer shrinkage in response to neoadjuvant therapy has
been shown to occur in 2 different ways: concentrically, and in a honey-
comb or buckshot pattern. MRI is the most reliable means of assessing both
the amount of residual tumor and the pattern of shrinkage.55 In patients
with a pathologic complete response or concentric tumor shrinkage, the
consideration of margin width does not differ from that in the primary sur-
gical setting. In patients with the honeycomb pattern, the determination
of the appropriate extent of resection is more problematic. Investigators
from the University of Texas MD Anderson Cancer Center have shown
that on multivariate analysis, a honeycomb pattern of response, residual
disease >2 cm, and clinical N2 or N3 disease were statistically significant
predictors of local recurrence after BCT.56 However, whether the higher
rate of local recurrence observed with the honeycomb pattern of shrinkage
can be reduced with more widely clear margins is uncertain because this
pattern of shrinkage may be a reflection of biologic properties of the tumor
which influence local recurrence just as they do in the primary surgical set-
ting. Evidence supporting this concept comes from a study of 149 patients

78
Principles and Techniques in Oncoplastic Breast Cancer Surgery
examining the impact of ER, PR, and HER2 status on local recurrence
after neoadjuvant therapy.57 The highest rate of local recurrence was seen
in the ER, PR, and HER2-negative group, just as was seen in studies exam-
ining the significance of these markers in the primary surgical setting.48,49
In the absence of definitive information on the benefit of more widely clear
margins after neoadjuvant therapy, it seems prudent to follow the current
joint guideline of the American College of Surgeons, American College
of Radiology, and College of American Pathologists which suggests that
if viable resident tumor is present scattered throughout the lumpectomy
specimen, even if it is not actually on the margin, re-excision should be
considered.58
5.2.4 Summary and Conclusions: Invasive Cancer
The failure of mastectomy, the most widely clear margin which can be
obtained in the breast to achieve rates of local control approaching 100%,
is clear evidence that the disease burden is not the only factor determining
local control. Evidence that margins, more widely clear than tumor not
touching ink, decrease local recurrence in patients receiving whole breast
RT is lacking, and the underlying biology of the tumor and the availability
of targeted therapy appear to be the major determinants of local control. In
recognition of the many factors impacting upon local control, a consensus
conference concluded that the only negative margin width that could be
endorsed for all patients with invasive carcinomas was tumor not touching
ink.59 This does not mean that in some circumstances a more widely clear
margin is not appropriate; however, it does mean that the routine use of
large surgical resection or mandatory re-excisions to obtain a more widely
clear margin in all patients should be abandoned.
5.3 Margin Width and Local Recurrence: DCIS
All of the caveats regarding the lack of a standardized approach to margin
assessment apply to DCIS as well as invasive cancer. In addition, because
DCIS is not clinically detectable in the overwhelming majority of cases, lim-
ited sampling of margin specimens is particularly likely to underestimate
the extent of DCIS. Three of the 4 randomized trials evaluating the benefit
of RT in DCIS used a margin definition of tumor not touching ink,60−62
while 20% patients in the SweDCIS study63 had positive or unknown mar-
gins. Just as in the case of invasive cancer, level 1 evidence from random-
ized trials on the effect of increasing margin width on local recurrence is
lacking. Low rates of local recurrence are seen with long-term follow-up in
studies using the tumor not touching ink definition of a negative margin.

Margins in Breast Cancer Surgery
79
In spite of the lack of a boost dose of RT in NSABP B17,61 the incidence
of in-breast recurrence at a median follow-up of 17.3 years was 20%. In
NSABP B24, this was reduced to 13.2% with the addition of tamoxifen,
even though positive margins were allowed in this study.61 In the United
Kingdom–Australia–New Zealand DCIS study (UK–ANZ), after a median
follow-up of 12.7 years, the incidence of in-breast recurrences was 10%.62
Dunne et al. performed a meta-analysis of 4,600 patients treated with
BCT to examine the question of margin width. Patients with negative mar-
gins were significantly less likely than those with positive margins to expe-
rience local recurrence (OR, 0.36; 95% CI, 0.27%–0.47%). When specific
margin widths were examined, patients with margins of 1 mm or less had a
higher rate of local recurrence than those with more widely clear margins,
but once a margin of 2 mm was obtained, no benefit for a more widely clear
margin was seen (Table 5.3).64 Of note, local recurrence was infrequent in
all of the negative margin groups. In a study of 994 women with DCIS
treated between 1985 and 2000, and identified through tumor registries in
Monroe County, New York and the Henry Ford Health System, Detroit,
Michigan, the multivariate relative risk of ipsilateral event-free survival for
patients with margins less than 2 mm was 1.39 (95% CI, 0.71–2.73) com-
pared to those with more widely clear margins.65 This translates to an
absolute difference of 1.6% at 5 years (95.6% versus 94.0%) and 3.2% at
10 years (90.5% versus 87.3%).
As previously discussed, DCIS may grow discontinuously in the ducts.
In a study of 60 mastectomy specimens with DCIS, Faverly et al.40 demon-
strated that 50% had a discontinuous growth pattern. This was seen in only
10% of high-grade DCIS lesions compared to 70% of well-differentiated
lesions and 55% of intermediate-grade lesions. These findings suggest that
margins more widely clear than tumor not touching ink are particularly rel-
evant for patients with low- and intermediate-grade DCIS. The presence of
residual calcifications at the lumpectomy site is a powerful predictor of the
presence of residual DCIS, and even when margins are negative, residual
suspicious calcifications are an indication for re-excision.66,67 Thus, margin
Table 5.3.
Impact of margin width on local recurrence in DCIS.
Negative margin width
Number
% In-breast recurrence
OR (95% CI)
Tumor not touching ink
914
9.4
2.6 (1.1–7.3)
1 mm
1239
10.4
2.9 (1.3–8.1)
2 mm
207
5.8
1.5 (0.5–5.0)
≥5 mm
154
3.9
1.0
CI, confidence interval; DCIS, ductal carcinoma in situ; OR, odds ratio.
Notes: Data from Dunne et al.64