local management_01

This section describes the multidisciplinary approach to the local
management of breast cancer by addressing the use of mastectomy,
conservative surgery (CS), and RT in a coordinated fashion, as well as
by considering the integration of local and systemic treatment.
Modified radical mastectomy is still the most common surgical
treatment for patients with invasive breast cancer in the United
States.(ref: 230,231) The term modified radical mastectomy is used to
describe a variety of surgical procedures, but all involve complete
removal of the breast, the underlying pectoral fascia, and some of the
axillary nodes. Whereas the modified radical mastectomy may not seem to
differ significantly from the radical mastectomy, it represents a major
departure from Halstedian principles of en bloc cancer surgery. The
switch to modified radical mastectomy occurred when it became
recognized that treatment failure after breast cancer surgery usually
is caused by the systemic dissemination of cancer cells before surgery,
rather than an inadequate operative procedure. In addition, by the
1970s, fewer patients with large tumors with fixation to the pectoral
muscle were being seen, making modified radical mastectomy feasible for
most women. Two prospective randomized trials demonstrated no
difference in survival between patients treated with modified radical
and radical mastectomy. These findings were confirmed in two
prospective randomized trials.(ref: 232,233) Perhaps the most
influential of the studies refuting the Halstedian concept was the
NSABP B-04 trial.(ref: 234) In this trial, clinically node-negative
patients were randomized to radical mastectomy, simple mastectomy and
nodal irradiation, or simple mastectomy with axillary observation and
delayed dissection if positive nodes developed. The failure of this
trial to demonstrate a difference in survival between groups was the
final proof that the Halstedian concept of breast cancer did not apply
to the majority of patients and was a landmark in our understanding of
the local therapy of breast cancer. Today, there are few, if any,
indications for radical mastectomy.
The strategy behind BCT is to remove the bulk of the tumor surgically
and to use moderate doses of radiation to eradicate any residual
cancer. The application of this strategy requires an understanding of
the extent and distribution of cancer in a breast with an apparently
localized tumor. This issue has been clarified as a result of the work
of Holland and coauthors. (ref: 235,236) In their initial study, (ref:
235) mastectomy specimens with unicentric tumors 4 cm or less in size
were evaluated using 5-mm sections, radiography of these thin slices,
and an average of 20 blocks per specimen for histologic evaluation.
Only 39% of specimens showed no evidence of cancer beyond the reference
tumor. In 20%, there was additional cancer, but this was confined to
within 2 cm of the reference tumor. Forty-one percent of cases had
residual cancer more than 2 cm from the reference tumor; of these, two-
thirds had pure intraductal carcinoma and one-third had mixed
intraductal and invasive carcinoma (Fig. 37.2_1). Local recurrence in
the breast occurs at or near the site of the primary tumor in most
cases,(ref: 237-240) emphasizing that this multifocal involvement is
biologically important. In a subsequent study, the amount of residual
intraductal carcinoma was evaluated.(ref: 236) Approximately 10% of
patients had prominent intraductal carcinoma (defined as a total of six
or more low-power fields of intraductal carcinoma) extending more than
2 cm from the reference tumor. These studies indicate that the extent
and amount of microscopic cancer in the vicinity of a primary tumor,
known as multifocality, is variable. These results imply that the
extent of surgical resection required in BCT varies from patient to
patient.
The published results of modern, prospective randomized clinical
trials comparing CS and RT and mastectomy have all shown equivalent
survival between the two treatment approaches, (ref: 241-247) and an
overview of all the trials (including an unpublished one) has
demonstrated comparable survival (ref: 248) (Fig. 37.2_2). These data
demonstrate that survival for most breast cancer patients is not
dependent on choice of local therapy. In addition to the results of
these trials, numerous reports from centers in Europe and North America
on the use of CS and RT have demonstrated high rates of local tumor
control with satisfactory cosmetic results. (ref: 249-253)
Despite the consistency of the evidence, the use of BCT in the United
States has shown relatively slow acceptance and considerable geographic
variation.(ref: 230,231) Studies indicate that fewer than 50% of women
with stage I and II breast carcinoma are treated with BCT. (ref:
230,231) The available data indicate that a minority of patients have
contraindications to BCT,(ref: 254,255) and that these are readily
identified with standard clinical tools, such as physical examination
and mammography including magnification views.(ref: 256) National
studies indicate that physicians continue to use inappropriate
selection criteria for BCT.(ref: 230)
The rates of recurrence in the breast at 7 to 18 years ranged from 7%
to 19% in the randomized studies using widely varying surgical and RT
techniques. (ref: 241-247) In the corresponding patients treated with
mastectomy, 4% to 14% of patients developed local recurrence,
emphasizing that mastectomy does not guarantee freedom from local
recurrence, even in women with clinical stage I and II breast
carcinoma. The nonrandomized studies with the longest follow-up
describe a persistent risk of recurrence in the breast through 20 years
of follow-up. (ref: 251-253,257) These results have been contrasted to
those seen after mastectomy, in which most local failures occur in the
first 3 years following surgery. The annual incidence rate for a
recurrence at or near the primary site is constant for years 2 through
7 after treatment, and then decreases to a low level by 10 years after
treatment. (ref: 253) In contrast, the annual incidence rate for
recurrence elsewhere in the breast increases slowly to a rate of
approximately 0.7% per year at 8 years and remains stable. (ref: 253)
Recurrences in the skin of the treated breast are a rare event
associated with a poor prognosis.(ref: 258) Whole breast irradiation is
effective at eradicating multicentric breast carcinoma, but it does not
prevent the subsequent development of new cancers.
A number of factors have been identified that influence the risk for
local recurrence after BCT. Young age has consistently been observed to
be associated with an increased risk of local recurrence after breast-
conserving surgery and RT. (ref: 259-263) However, young age has
similarly been associated with a worse outcome after mastectomy. (ref:
264,265) In young women with a family history suggestive of an
inherited breast cancer susceptibility, BCT is associated with a higher
rate of opposite breast cancer compared with young women without such a
family history. (ref: 266) This is consistent with the findings of an
increased risk of opposite breast cancer in young patients with
mutations undergoing mastectomy. (ref: 31,267) The rate of local
recurrence in young patients with a positive family history is, if
anything, lower than in patients with a negative family history. This
might be explained by the findings linking BRCA1 and 2 with radiation
repair genes, (ref: 268,269) or by a greater likelihood of localized
(extensive intraductal component-negative) cancers in patients with
mutations compared with patients without mutations.(ref: 270) However,
patients with mutations appear to be at risk for late new primaries in
the treated breast.(ref: 271) Of note, patients with a mutation do not
appear to be at an increased risk for adverse effects from RT. (ref:
272,273) Thus, BCT appears to be an acceptable option for patients with
a suspected or known mutation, although these patients need to be
apprised of the increased risk of a second breast cancer, either in the
opposite or, over time, in the treated breast. Many of these patients,
particularly those with favorable presentations, elect bilateral
mastectomy. A modeling study suggests that bilateral mastectomy may be
associated with a modest gain in survival. (ref: 127)
An extensive intraductal component has been shown to be an important
risk factor for local recurrence when margins of resection are not
evaluated.(ref: 274) An extensive intraductal component has been found
to be a marker for a large residual tumor burden in the involved
quadrant of the breast (ref: 236,275) such that moderate-dose RT is not
able to eradicate it. In such patients, a larger breast resection is
commonly required to ensure adequate removal. Results have shown that
the microscopic margins of resection are the major selection factor for
BCT (Tables 37.2_11 and 37.2_12). Patients with negative margins of
excision (typically defined as the absence of either invasive or ductal
in situ disease directly at an inked surface) have generally been
observed to have low rates of local recurrence following treatment with
CS and RT. (ref: 276-285) In particular, patients with an extensive
intraductal component, but with negative inked margins of excision, are
not at an increased risk of local recurrence.(ref: 279-281,286) The
outcome of patients with close margins of excision has been less clear.
In part, this reflects variability in the definition of close margins
and, perhaps, the effect of institutional policies calling for
escalated radiation doses based on the proximity of cancer cells to the
margin of resection. In the Joint Center for Radiation Therapy (JCRT)
experience shown in Table 37.2_12, there was no significant
difference in recurrence rates between patients with close margins
(less than or equal to 1 mm) compared with patients with margins
greater than 1 mm using similar doses. (ref: 279) Some studies have
suggested a high rate of local recurrence at 10 years in patients with
close margins; however, the number of patients in these series and the
actual follow-up time is limited. (ref: 278,281)
Long-term data on the use of BCT in patients with positive margins
are more limited. In most analyses, positive margins have been
associated with a high risk of breast cancer recurrence. (ref: 276-
280,283-285) At the JCRT, patients with positive margins had a
considerably higher risk of breast cancer recurrence than patients with
negative margins. (ref: 279) The 8-year crude rate of breast recurrence
was 18% for patients with positive margins. However, patients with
focally positive margins (any invasive or in situ ductal carcinoma at
the margin in three or fewer low-power microscopic fields) had a 14%
rate of recurrence compared with a 27% rate in patients with greater
than focally positive margins. These data suggest that patients with
focally positive margins can be considered for BCT. As discussed in
this section, the use of adjuvant systemic therapy results in a large
reduction in local recurrence in patients treated with CS and RT. In
the JCRT series, among the 45 patients with focally positive margins
who received adjuvant systemic therapy, the 8-year local recurrence
rate was 8% (95% confidence interval, 1% to 18%). (ref: 279) Additional
experience is needed to confirm this finding. Patients with more than
focally positive margins require more surgery given the significantly
higher rate of breast cancer recurrence.
The use of adjuvant systemic therapy is an important factor
associated with recurrence in the breast when used in conjunction with
CS and RT. This is most clearly demonstrated in three randomized
clinical trials. In the NSABP B-13 trial, node-negative, ER-negative
patients were randomized to chemotherapy or to a no-treatment control
group. Among the 235 patients treated with CS and RT, the 8-year rate
of recurrence in the ipsilateral breast was 13.4% without chemotherapy
and only 2.6% with chemotherapy. (ref: 287) Similar results are seen
with adjuvant tamoxifen. In NSABP trial B-14, node-negative, ER-
positive patients were randomized to tamoxifen or to a placebo. Among
the 1062 patients treated with CS and RT, the 10-year rate of
recurrence in the ipsilateral breast was 14.7% without tamoxifen and
only 4.3% with tamoxifen. (ref: 288) A similar result was seen in the
Stockholm Breast Cancer Study Group among node-negative patients
randomized to tamoxifen or to a placebo. (ref: 289) Among the 432
patients treated with CS and RT, the 10-year rate of recurrence in the
ipsilateral breast was 12% without tamoxifen and only 3% with
tamoxifen.

Guidelines for Patient Selection

Based on the extensive information available from prospective and
retrospective studies, there is a general consensus on the criteria for
patient selection for the use of BCT. It is now established that, in
most cases, BCT results in a cosmetically satisfactory breast and that
it provides survival rates equivalent to those seen after mastectomy.
The American College of Surgeons, the American College of Radiology,
the College of American Pathologists, and the Society of Surgical
Oncology have jointly provided standards of care for BCT and most
recently published their report in 1998. (ref: 290) Key portions of
this report are summarized here and additional comments are provided in
parentheses.

Contraindications for Breast-Conservation Treatment with Radiation
Therapy

Absolute Contraindications

- Women with two or more primary tumors in separate quadrants of
the breast or with diffuse malignant-appearing microcalcifications are
not considered candidates for breast-conservation treatment.
- A history of previous therapeutic irradiation to the breast
region that, combined with the proposed treatment, would result in an
excessively high total radiation dose to a significant volume is
another absolute contraindication.
- Pregnancy is an absolute contraindication to the use of breast
irradiation. However, in many cases, it may be possible to perform
breast-conserving surgery in the third trimester and to treat the
patient with irradiation after delivery.
- Finally, persistent positive margins after reasonable surgical
attempts absolutely contraindicate BCT with radiation. The importance
of a single focally positive microscopic margin needs further study and
may not be an absolute contraindication (see updated results from the
JCRT in Local Management of Invasive Breast Cancer, earlier in this
chapter).

Relative Contraindications

- A history of collagen vascular disease is a relative
contraindication to BCT because published reports indicate that such
patients tolerate irradiation poorly. (ref: 291) Most radiation
oncologists will not treat patients with scleroderma or active lupus
erythematosus, considering either an absolute contraindication. In
contrast, rheumatoid arthritis is not a contraindication. (ref: 292)
- Patients with multiple gross tumors in the same quadrant and
indeterminate calcifications must be carefully assessed for suitability
because studies in this area are not definitive.
- Tumor size is not an absolute contraindication to BCT, although
few reports have been published about treating patients with tumors
larger than 4 to 5 cm. However, a relative contraindication is the
presence of a large tumor in a small breast in which an adequate
resection would result in significant cosmetic alteration.
- Breast size can be a relative contraindication. Women with large
or pendulous breasts can be treated by irradiation if reproducibility
of patient setup can be ensured and it is technically possible to
obtain adequate dose homogeneity.

Nonmitigating Factors

- The presence of clinical or pathologic involvement in axillary
nodes should not prevent the treatment.
- Concern about not being able to detect a recurrence is not a
contraindication. The changes associated with recurrence can usually be
detected at an early stage by physical examination and mammography.
- The delivery of irradiation to the breast does not result in a
meaningful risk of second tumors in the treated area or in the
untreated area.
- Tumor location is not a factor in the choice of treatment. Tumors
in a superficial subareolar location occasionally may require the
resection of the nipple-areolar complex so that negative margins can be
achieved, but this does not affect outcome. The patient and her
physician need to assess whether such a resection is preferable to
mastectomy.
- A family history of breast cancer is not a contraindication to
breast conservation. Little is known about the risk of breast
recurrence in patients with hereditary breast cancer, but currently
this is not a contraindication to BCT. (However, such patients should
be apprised of their increased risk of a second breast cancer.)
- A high risk of systemic relapse is not a contraindication for
breast conservation, but is a determinant of the need for adjuvant
therapy.

Conservative Surgery without Radiation Therapy

An unresolved question is whether RT is necessary in all patients with
invasive breast cancer after CS. Six randomized clinical trials with
published results have compared CS alone with CS and RT in patients
with early-stage breast cancer. (ref: 241,242,293-296) These trials
vary with regard to patient selection, the details of the surgery and
RT, the use of adjuvant systemic therapy, and the length of follow-up.
The results of these various trials are shown in Table 37.2_13. These
trials all show a large reduction in the rate of local recurrence after
RT, with an average crude rate of reduction of approximately 75%
(range, 63% to 89%). None of the six trials shows a significant
survival benefit for RT; however, in the trials with published data,
the survival rate is slightly better for irradiated patients than for
nonirradiated patients. A large trial (or perhaps a metaanalysis of
multiple smaller trials) is necessary to detect a small, but clinically
significant difference in survival, if it in fact exists.
Attempts have been made to identify a subgroup of patients (based on
various clinical and histologic features) that has a low risk of local
recurrence after CS alone. It was not possible to identify such a
subgroup within the Ontario and NSABP randomized trials. Local
recurrence rates are generally lower in trials using more extensive
surgery than in those using lumpectomy and in older patients than in
younger patients. The JCRT attempted to identify such a subgroup in a
prospective single-arm trial in which patients with favorable disease
were offered the option of CS alone. The criteria for entry onto this
protocol were tumor size of 2 cm or less, histologically negative
axillary nodes, absence of both lymphatic vessel invasion and an
extensive intraductal component in the cancer, and no cancer cells
visualized within 1 cm of inked margins. (ref: 297) All but one patient
had a negative reexcision. This trial was stopped shortly before
reaching its accrual goal of 90 patients because of stopping rules
ensuring against an excessively high local recurrence rate. The latest
analysis includes the results in 81 patients. The median age of
patients in this trial was 66 years, and median pathologic size of the
cancers was 9 mm. With a median follow-up of 92 months, 19 of the
patients have developed a recurrence in the ipsilateral breast, for a
crude local recurrence rate of 23%. Based on the results of this
prospective study, it was concluded that, even in a highly selected
group of breast cancer patients, there is a substantial risk of early
local recurrence after treatment with wide excision alone.
The use of adjuvant systemic therapy substantially reduces the rate
of local recurrence in patients treated with CS and RT, (ref: 287-289)
but does not seem to reduce greatly the rate of local recurrence after
CS alone. There are no published trials directly comparing CS with and
without either chemotherapy or tamoxifen. Information on this is
available from indirect comparisons within randomized clinical trials
for both adjuvant chemotherapy and tamoxifen. In the NSABP trial B-06,
an indirect comparison of the effect of adjuvant chemotherapy can be
made. Node-positive patients treated with lumpectomy and adjuvant
chemotherapy but without RT had a 12-year rate of recurrence in the
breast of 41% compared with only 5% for node-positive patients treated
with lumpectomy, RT, and chemotherapy (P <.001). (ref: 241) In
comparison, node-negative patients treated with lumpectomy without RT
had a 12-year rate of recurrence in the breast of 32% compared with 12%
for node-negative patients treated with lumpectomy with RT. A similar
observation, suggesting that systemic therapy further decreases the
rate of local recurrence when combined with RT, but not in its absence,
is also seen in indirect comparisons within the Milan trials. (ref:
242) In the Scottish trial, patients with ER-negative cancers were
treated with adjuvant cyclophosphamide, methotrexate, and 5-
fluorouracil (CMF) chemotherapy. With a median follow-up of
approximately 5.7 years, the crude rate of local regional recurrence
was 44% among patients treated with CS, but without RT, compared with
only 14% among patients treated with RT.(ref: 295)
There is particular interest in avoiding RT in older patients. It is
often less convenient for such patients to receive RT, and their local
recurrence rate appears lower after CS alone compared with younger
patients.(ref: 298-300) The results of retrospective studies of CS
alone with or without adjuvant tamoxifen have shown variable results.
(ref: 301-302) The Cancer and Leukemia Group B (CALGB) and other groups
in North America have completed a prospective randomized clinical trial
testing the value of RT in older breast cancer patients treated by CS
and tamoxifen; at this time there are no results from the trial. In
NSABP B-21, women with tumors smaller than 1 cm with negative axillary
lymph nodes were randomized to tamoxifen alone, breast irradiation
alone, or breast irradiation plus tamoxifen. With an average follow-up
of 73 months, 24.4% of women in the tamoxifen-only arm had an
ipsilateral recurrence, compared to 11.7% of women who received breast
irradiation plus tamoxifen. The difference was highly statistically
significant. Based on these results, the investigators thought it
unlikely that tamoxifen could be substituted for radiation in this
patient population.(ref: 300a)
In conclusion, the use of breast irradiation after CS is associated
with a large reduction in the rate of local recurrence. The available
data from the randomized trials do not show a survival benefit;
however, none of the available trials has the statistical power to
eliminate a small survival difference. A subset at low risk of local
recurrence following CS has not been clearly identified, and RT is
currently considered standard. The addition of adjuvant systemic
therapy to CS alone has not been demonstrated to decrease local
recurrence. In elderly patients, particularly those with significant
comorbidity, RT is commonly omitted because of the practical
difficulties of delivering such therapy in this group of patients.

local management

Local Management of Invasive Breast Cancer

This section describes the multidisciplinary approach to the local
management of breast cancer by addressing the use of mastectomy,
conservative surgery (CS), and RT in a coordinated fashion, as well as
by considering the integration of local and systemic treatment.
Modified radical mastectomy is still the most common surgical
treatment for patients with invasive breast cancer in the United
States.[ref: 230,231] The term modified radical mastectomy is used to
describe a variety of surgical procedures, but all involve complete
removal of the breast, the underlying pectoral fascia, and some of the
axillary nodes. Whereas the modified radical mastectomy may not seem to
differ significantly from the radical mastectomy, it represents a major
departure from Halstedian principles of en bloc cancer surgery. The
switch to modified radical mastectomy occurred when it became
recognized that treatment failure after breast cancer surgery usually
is caused by the systemic dissemination of cancer cells before surgery,
rather than an inadequate operative procedure. In addition, by the
1970s, fewer patients with large tumors with fixation to the pectoral
muscle were being seen, making modified radical mastectomy feasible for
most women. Two prospective randomized trials demonstrated no
difference in survival between patients treated with modified radical
and radical mastectomy. These findings were confirmed in two
prospective randomized trials.[ref: 232,233] Perhaps the most
influential of the studies refuting the Halstedian concept was the
NSABP B-04 trial.[ref: 234] In this trial, clinically node-negative
patients were randomized to radical mastectomy, simple mastectomy and
nodal irradiation, or simple mastectomy with axillary observation and
delayed dissection if positive nodes developed. The failure of this
trial to demonstrate a difference in survival between groups was the
final proof that the Halstedian concept of breast cancer did not apply
to the majority of patients and was a landmark in our understanding of
the local therapy of breast cancer. Today, there are few, if any,
indications for radical mastectomy.
The strategy behind BCT is to remove the bulk of the tumor surgically
and to use moderate doses of radiation to eradicate any residual
cancer. The application of this strategy requires an understanding of
the extent and distribution of cancer in a breast with an apparently
localized tumor. This issue has been clarified as a result of the work
of Holland and coauthors. [ref: 235,236] In their initial study, [ref:
235] mastectomy specimens with unicentric tumors 4 cm or less in size
were evaluated using 5-mm sections, radiography of these thin slices,
and an average of 20 blocks per specimen for histologic evaluation.
Only 39% of specimens showed no evidence of cancer beyond the reference
tumor. In 20%, there was additional cancer, but this was confined to
within 2 cm of the reference tumor. Forty-one percent of cases had
residual cancer more than 2 cm from the reference tumor; of these, two-
thirds had pure intraductal carcinoma and one-third had mixed
intraductal and invasive carcinoma (Fig. 37.2_1). Local recurrence in
the breast occurs at or near the site of the primary tumor in most
cases,[ref: 237-240] emphasizing that this multifocal involvement is
biologically important. In a subsequent study, the amount of residual
intraductal carcinoma was evaluated.[ref: 236] Approximately 10% of
patients had prominent intraductal carcinoma (defined as a total of six
or more low-power fields of intraductal carcinoma) extending more than
2 cm from the reference tumor. These studies indicate that the extent
and amount of microscopic cancer in the vicinity of a primary tumor,
known as multifocality, is variable. These results imply that the
extent of surgical resection required in BCT varies from patient to
patient.
The published results of modern, prospective randomized clinical
trials comparing CS and RT and mastectomy have all shown equivalent
survival between the two treatment approaches, [ref: 241-247] and an
overview of all the trials (including an unpublished one) has
demonstrated comparable survival [ref: 248] (Fig. 37.2_2). These data
demonstrate that survival for most breast cancer patients is not
dependent on choice of local therapy. In addition to the results of
these trials, numerous reports from centers in Europe and North America
on the use of CS and RT have demonstrated high rates of local tumor
control with satisfactory cosmetic results. [ref: 249-253]
Despite the consistency of the evidence, the use of BCT in the United
States has shown relatively slow acceptance and considerable geographic
variation.[ref: 230,231] Studies indicate that fewer than 50% of women
with stage I and II breast carcinoma are treated with BCT. [ref:
230,231] The available data indicate that a minority of patients have
contraindications to BCT,[ref: 254,255] and that these are readily
identified with standard clinical tools, such as physical examination
and mammography including magnification views.[ref: 256] National
studies indicate that physicians continue to use inappropriate
selection criteria for BCT.[ref: 230]
The rates of recurrence in the breast at 7 to 18 years ranged from 7%
to 19% in the randomized studies using widely varying surgical and RT
techniques. [ref: 241-247] In the corresponding patients treated with
mastectomy, 4% to 14% of patients developed local recurrence,
emphasizing that mastectomy does not guarantee freedom from local
recurrence, even in women with clinical stage I and II breast
carcinoma. The nonrandomized studies with the longest follow-up
describe a persistent risk of recurrence in the breast through 20 years
of follow-up. [ref: 251-253,257] These results have been contrasted to
those seen after mastectomy, in which most local failures occur in the
first 3 years following surgery. The annual incidence rate for a
recurrence at or near the primary site is constant for years 2 through
7 after treatment, and then decreases to a low level by 10 years after
treatment. [ref: 253] In contrast, the annual incidence rate for
recurrence elsewhere in the breast increases slowly to a rate of
approximately 0.7% per year at 8 years and remains stable. [ref: 253]
Recurrences in the skin of the treated breast are a rare event
associated with a poor prognosis.[ref: 258] Whole breast irradiation is
effective at eradicating multicentric breast carcinoma, but it does not
prevent the subsequent development of new cancers.
A number of factors have been identified that influence the risk for
local recurrence after BCT. Young age has consistently been observed to
be associated with an increased risk of local recurrence after breast-
conserving surgery and RT. [ref: 259-263] However, young age has
similarly been associated with a worse outcome after mastectomy. [ref:
264,265] In young women with a family history suggestive of an
inherited breast cancer susceptibility, BCT is associated with a higher
rate of opposite breast cancer compared with young women without such a
family history. [ref: 266] This is consistent with the findings of an
increased risk of opposite breast cancer in young patients with
mutations undergoing mastectomy. [ref: 31,267] The rate of local
recurrence in young patients with a positive family history is, if
anything, lower than in patients with a negative family history. This
might be explained by the findings linking BRCA1 and 2 with radiation
repair genes, [ref: 268,269] or by a greater likelihood of localized
(extensive intraductal component-negative) cancers in patients with
mutations compared with patients without mutations.[ref: 270] However,
patients with mutations appear to be at risk for late new primaries in
the treated breast.[ref: 271] Of note, patients with a mutation do not
appear to be at an increased risk for adverse effects from RT. [ref:
272,273] Thus, BCT appears to be an acceptable option for patients with
a suspected or known mutation, although these patients need to be
apprised of the increased risk of a second breast cancer, either in the
opposite or, over time, in the treated breast. Many of these patients,
particularly those with favorable presentations, elect bilateral
mastectomy. A modeling study suggests that bilateral mastectomy may be
associated with a modest gain in survival. [ref: 127]
An extensive intraductal component has been shown to be an important
risk factor for local recurrence when margins of resection are not
evaluated.[ref: 274] An extensive intraductal component has been found
to be a marker for a large residual tumor burden in the involved
quadrant of the breast [ref: 236,275] such that moderate-dose RT is not
able to eradicate it. In such patients, a larger breast resection is
commonly required to ensure adequate removal. Results have shown that
the microscopic margins of resection are the major selection factor for
BCT (Tables 37.2_11 and 37.2_12). Patients with negative margins of
excision (typically defined as the absence of either invasive or ductal
in situ disease directly at an inked surface) have generally been
observed to have low rates of local recurrence following treatment with
CS and RT. [ref: 276-285] In particular, patients with an extensive
intraductal component, but with negative inked margins of excision, are
not at an increased risk of local recurrence.[ref: 279-281,286] The
outcome of patients with close margins of excision has been less clear.
In part, this reflects variability in the definition of close margins
and, perhaps, the effect of institutional policies calling for
escalated radiation doses based on the proximity of cancer cells to the
margin of resection. In the Joint Center for Radiation Therapy (JCRT)
experience shown in Table 37.2_12, there was no significant
difference in recurrence rates between patients with close margins
(less than or equal to 1 mm) compared with patients with margins
greater than 1 mm using similar doses. [ref: 279] Some studies have
suggested a high rate of local recurrence at 10 years in patients with
close margins; however, the number of patients in these series and the
actual follow-up time is limited. [ref: 278,281]
Long-term data on the use of BCT in patients with positive margins
are more limited. In most analyses, positive margins have been
associated with a high risk of breast cancer recurrence. [ref: 276-
280,283-285] At the JCRT, patients with positive margins had a
considerably higher risk of breast cancer recurrence than patients with
negative margins. [ref: 279] The 8-year crude rate of breast recurrence
was 18% for patients with positive margins. However, patients with
focally positive margins (any invasive or in situ ductal carcinoma at
the margin in three or fewer low-power microscopic fields) had a 14%
rate of recurrence compared with a 27% rate in patients with greater
than focally positive margins. These data suggest that patients with
focally positive margins can be considered for BCT. As discussed in
this section, the use of adjuvant systemic therapy results in a large
reduction in local recurrence in patients treated with CS and RT. In
the JCRT series, among the 45 patients with focally positive margins
who received adjuvant systemic therapy, the 8-year local recurrence
rate was 8% (95% confidence interval, 1% to 18%). [ref: 279] Additional
experience is needed to confirm this finding. Patients with more than
focally positive margins require more surgery given the significantly
higher rate of breast cancer recurrence.
The use of adjuvant systemic therapy is an important factor
associated with recurrence in the breast when used in conjunction with
CS and RT. This is most clearly demonstrated in three randomized
clinical trials. In the NSABP B-13 trial, node-negative, ER-negative
patients were randomized to chemotherapy or to a no-treatment control
group. Among the 235 patients treated with CS and RT, the 8-year rate
of recurrence in the ipsilateral breast was 13.4% without chemotherapy
and only 2.6% with chemotherapy. [ref: 287] Similar results are seen
with adjuvant tamoxifen. In NSABP trial B-14, node-negative, ER-
positive patients were randomized to tamoxifen or to a placebo. Among
the 1062 patients treated with CS and RT, the 10-year rate of
recurrence in the ipsilateral breast was 14.7% without tamoxifen and
only 4.3% with tamoxifen. [ref: 288] A similar result was seen in the
Stockholm Breast Cancer Study Group among node-negative patients
randomized to tamoxifen or to a placebo. [ref: 289] Among the 432
patients treated with CS and RT, the 10-year rate of recurrence in the
ipsilateral breast was 12% without tamoxifen and only 3% with
tamoxifen.

Guidelines for Patient Selection

Based on the extensive information available from prospective and
retrospective studies, there is a general consensus on the criteria for
patient selection for the use of BCT. It is now established that, in
most cases, BCT results in a cosmetically satisfactory breast and that
it provides survival rates equivalent to those seen after mastectomy.
The American College of Surgeons, the American College of Radiology,
the College of American Pathologists, and the Society of Surgical
Oncology have jointly provided standards of care for BCT and most
recently published their report in 1998. [ref: 290] Key portions of
this report are summarized here and additional comments are provided in
parentheses.

Contraindications for Breast-Conservation Treatment with Radiation
Therapy

Absolute Contraindications

- Women with two or more primary tumors in separate quadrants of
the breast or with diffuse malignant-appearing microcalcifications are
not considered candidates for breast-conservation treatment.
- A history of previous therapeutic irradiation to the breast
region that, combined with the proposed treatment, would result in an
excessively high total radiation dose to a significant volume is
another absolute contraindication.
- Pregnancy is an absolute contraindication to the use of breast
irradiation. However, in many cases, it may be possible to perform
breast-conserving surgery in the third trimester and to treat the
patient with irradiation after delivery.
- Finally, persistent positive margins after reasonable surgical
attempts absolutely contraindicate BCT with radiation. The importance
of a single focally positive microscopic margin needs further study and
may not be an absolute contraindication (see updated results from the
JCRT in Local Management of Invasive Breast Cancer, earlier in this
chapter).

Relative Contraindications

- A history of collagen vascular disease is a relative
contraindication to BCT because published reports indicate that such
patients tolerate irradiation poorly. [ref: 291] Most radiation
oncologists will not treat patients with scleroderma or active lupus
erythematosus, considering either an absolute contraindication. In
contrast, rheumatoid arthritis is not a contraindication. [ref: 292]
- Patients with multiple gross tumors in the same quadrant and
indeterminate calcifications must be carefully assessed for suitability
because studies in this area are not definitive.
- Tumor size is not an absolute contraindication to BCT, although
few reports have been published about treating patients with tumors
larger than 4 to 5 cm. However, a relative contraindication is the
presence of a large tumor in a small breast in which an adequate
resection would result in significant cosmetic alteration.
- Breast size can be a relative contraindication. Women with large
or pendulous breasts can be treated by irradiation if reproducibility
of patient setup can be ensured and it is technically possible to
obtain adequate dose homogeneity.

Nonmitigating Factors

- The presence of clinical or pathologic involvement in axillary
nodes should not prevent the treatment.
- Concern about not being able to detect a recurrence is not a
contraindication. The changes associated with recurrence can usually be
detected at an early stage by physical examination and mammography.
- The delivery of irradiation to the breast does not result in a
meaningful risk of second tumors in the treated area or in the
untreated area.
- Tumor location is not a factor in the choice of treatment. Tumors
in a superficial subareolar location occasionally may require the
resection of the nipple-areolar complex so that negative margins can be
achieved, but this does not affect outcome. The patient and her
physician need to assess whether such a resection is preferable to
mastectomy.
- A family history of breast cancer is not a contraindication to
breast conservation. Little is known about the risk of breast
recurrence in patients with hereditary breast cancer, but currently
this is not a contraindication to BCT. (However, such patients should
be apprised of their increased risk of a second breast cancer.)
- A high risk of systemic relapse is not a contraindication for
breast conservation, but is a determinant of the need for adjuvant
therapy.

Conservative Surgery without Radiation Therapy

An unresolved question is whether RT is necessary in all patients with
invasive breast cancer after CS. Six randomized clinical trials with
published results have compared CS alone with CS and RT in patients
with early-stage breast cancer. [ref: 241,242,293-296] These trials
vary with regard to patient selection, the details of the surgery and
RT, the use of adjuvant systemic therapy, and the length of follow-up.
The results of these various trials are shown in Table 37.2_13. These
trials all show a large reduction in the rate of local recurrence after
RT, with an average crude rate of reduction of approximately 75%
(range, 63% to 89%). None of the six trials shows a significant
survival benefit for RT; however, in the trials with published data,
the survival rate is slightly better for irradiated patients than for
nonirradiated patients. A large trial (or perhaps a metaanalysis of
multiple smaller trials) is necessary to detect a small, but clinically
significant difference in survival, if it in fact exists.
Attempts have been made to identify a subgroup of patients (based on
various clinical and histologic features) that has a low risk of local
recurrence after CS alone. It was not possible to identify such a
subgroup within the Ontario and NSABP randomized trials. Local
recurrence rates are generally lower in trials using more extensive
surgery than in those using lumpectomy and in older patients than in
younger patients. The JCRT attempted to identify such a subgroup in a
prospective single-arm trial in which patients with favorable disease
were offered the option of CS alone. The criteria for entry onto this
protocol were tumor size of 2 cm or less, histologically negative
axillary nodes, absence of both lymphatic vessel invasion and an
extensive intraductal component in the cancer, and no cancer cells
visualized within 1 cm of inked margins. [ref: 297] All but one patient
had a negative reexcision. This trial was stopped shortly before
reaching its accrual goal of 90 patients because of stopping rules
ensuring against an excessively high local recurrence rate. The latest
analysis includes the results in 81 patients. The median age of
patients in this trial was 66 years, and median pathologic size of the
cancers was 9 mm. With a median follow-up of 92 months, 19 of the
patients have developed a recurrence in the ipsilateral breast, for a
crude local recurrence rate of 23%. Based on the results of this
prospective study, it was concluded that, even in a highly selected
group of breast cancer patients, there is a substantial risk of early
local recurrence after treatment with wide excision alone.
The use of adjuvant systemic therapy substantially reduces the rate
of local recurrence in patients treated with CS and RT, [ref: 287-289]
but does not seem to reduce greatly the rate of local recurrence after
CS alone. There are no published trials directly comparing CS with and
without either chemotherapy or tamoxifen. Information on this is
available from indirect comparisons within randomized clinical trials
for both adjuvant chemotherapy and tamoxifen. In the NSABP trial B-06,
an indirect comparison of the effect of adjuvant chemotherapy can be
made. Node-positive patients treated with lumpectomy and adjuvant
chemotherapy but without RT had a 12-year rate of recurrence in the
breast of 41% compared with only 5% for node-positive patients treated
with lumpectomy, RT, and chemotherapy (P <.001). [ref: 241] In
comparison, node-negative patients treated with lumpectomy without RT
had a 12-year rate of recurrence in the breast of 32% compared with 12%
for node-negative patients treated with lumpectomy with RT. A similar
observation, suggesting that systemic therapy further decreases the
rate of local recurrence when combined with RT, but not in its absence,
is also seen in indirect comparisons within the Milan trials. [ref:
242] In the Scottish trial, patients with ER-negative cancers were
treated with adjuvant cyclophosphamide, methotrexate, and 5-
fluorouracil (CMF) chemotherapy. With a median follow-up of
approximately 5.7 years, the crude rate of local regional recurrence
was 44% among patients treated with CS, but without RT, compared with
only 14% among patients treated with RT.[ref: 295]
There is particular interest in avoiding RT in older patients. It is
often less convenient for such patients to receive RT, and their local
recurrence rate appears lower after CS alone compared with younger
patients.[ref: 298-300] The results of retrospective studies of CS
alone with or without adjuvant tamoxifen have shown variable results.
[ref: 301-302] The Cancer and Leukemia Group B (CALGB) and other groups
in North America have completed a prospective randomized clinical trial
testing the value of RT in older breast cancer patients treated by CS
and tamoxifen; at this time there are no results from the trial. In
NSABP B-21, women with tumors smaller than 1 cm with negative axillary
lymph nodes were randomized to tamoxifen alone, breast irradiation
alone, or breast irradiation plus tamoxifen. With an average follow-up
of 73 months, 24.4% of women in the tamoxifen-only arm had an
ipsilateral recurrence, compared to 11.7% of women who received breast
irradiation plus tamoxifen. The difference was highly statistically
significant. Based on these results, the investigators thought it
unlikely that tamoxifen could be substituted for radiation in this
patient population.[ref: 300a]
In conclusion, the use of breast irradiation after CS is associated
with a large reduction in the rate of local recurrence. The available
data from the randomized trials do not show a survival benefit;
however, none of the available trials has the statistical power to
eliminate a small survival difference. A subset at low risk of local
recurrence following CS has not been clearly identified, and RT is
currently considered standard. The addition of adjuvant systemic
therapy to CS alone has not been demonstrated to decrease local
recurrence. In elderly patients, particularly those with significant
comorbidity, RT is commonly omitted because of the practical
difficulties of delivering such therapy in this group of patients.

Preoperative Chemotherapy

The successful application of preoperative chemotherapy in locally
advanced breast cancer has led to a number of studies in patients with
stage I and II breast carcinoma to determine if the use of preoperative
chemotherapy would allow breast conservation in patients who would
otherwise be treated with mastectomy. Early studies demonstrated a high
response rate to preoperative chemotherapy with conversion of patients
initially thought to be unsuitable for BCT to be considered reasonable
candidates for breast conservation. [ref: 303,304] In addition to
increasing the number of patients who can undergo BCT, a second goal of
preoperative therapy is to improve survival. The available randomized
trials suggest that the use of preoperative chemotherapy does reduce
the use of mastectomy, but does not improve survival.[ref: 305-307] Of
note, however, in the NSABP trial B-18, [ref: 306] an analysis of
breast recurrence rates among patients initially eligible for
lumpectomy and those who were eligible only after down-staging by
chemotherapy demonstrated a local failure rate of 6.9% in those thought
to be candidates for lumpectomy before chemotherapy compared with 14.5%
in those who required down-staging (P = .04). Similar findings were
noted in a trial from France. [ref: 308] Further experience is
therefore needed to identify down-staged patients who can be
effectively managed by BCT. A major practical problem with the use of
preoperative chemotherapy to increase rates of BCT is the determination
of the extent of residual viable tumor that must be resected. The
clinical assessment of response is relatively inaccurate using clinical
examination and mammography. Given this, we approach these patients by
initially resecting any clinically or mammographically abnormal tissue.
If viable tumor is present throughout the specimen, a reexcision is
carried out even if the initial margins are negative. If further viable
tumor is present in the reexcised specimen, a reevaluation of the
patient's suitability for BCT is undertaken. Marking the extent of the
tumor before chemotherapy with stereotactically placed clips or skin
tattoos is useful for determining the tumor location in patients who
have a complete clinical response and may aid in assessing the need for
resection of residual abnormalities in patients with a partial
response.
The definitive role of neoadjuvant therapy in operable breast cancer
remains undefined. There appears to be no rationale, outside of a
clinical trial, for its routine use in patients who are suitable
candidates for BCT. Initial chemotherapy is appropriate when a large
tumor in a small breast would necessitate mastectomy and the patient
desires BCT. However, in the study of Morrow et al., [ref: 254] this
contraindication to BCT was present in only 6% of 336 patients with
stage I and stage II carcinoma. The potential for a higher risk of
breast recurrence should be discussed with the patient, and the
pathology carefully reviewed before deciding that the patient is a
suitable candidate for BCT.

Technique and Complications of Breast-Conserving Surgery

The goal of breast-conserving surgery is to minimize the risk of local
recurrence while leaving the patient with a cosmetically acceptable
breast. The most common form of breast-conserving surgery used in the
United States is referred to as lumpectomy. The surgical technique of
lumpectomy differs from that used for mastectomy in that lumpectomy is
not an en bloc cancer operation. Quadrantectomy is another type of
breast-conserving surgery that is designed to remove an anatomic
segment of breast tissue and frequently includes removal of the
overlying skin and underlying pectoral fascia. Because excision of a
large amount of breast tissue is the major factor responsible for a
poor cosmetic outcome after BCT, lumpectomy is considered the
appropriate initial surgical approach in the United States. Other
surgical factors that influence the cosmetic appearance are the size
and placement of the incision, the management of the lumpectomy cavity,
and the extent of axillary dissection.
A number of technical aspects of lumpectomy are worth emphasizing. In
general, the incision should be placed directly over the area of the
tumor. This is true even when a biopsy is performed for a
mammographically detected lesion. In the upper part of the breast,
incisions should be curvilinear or transverse and follow the natural
skin creases (Langer's lines) of the breast. In the lower part of the
breast, the choice of a curvilinear or radial incision depends on the
contour of the patient's breast, the distance from the skin to the
tumor, and the amount of breast tissue to be resected. It is not
necessary to remove skin (except for superficial tumors) or to remove
needle tracks from core-needle biopsies or FNAs. Preservation of the
subcutaneous fat and the avoidance of thin skin flaps is also important
in maintaining normal breast contour. Raising flaps is necessary only
to allow access to the tumor. Meticulous hemostasis is important
because a large hematoma distorts the appearance of the breast and
makes reexcision and follow-up evaluation more difficult. The presence
of a postbiopsy hematoma, however, is not a contraindication to BCT. It
is best to avoid reapproximation of the breast tissue since this can
result in distortion of the breast contour, which may not be apparent
with the patient supine on the operating table. The best cosmetic
results usually are obtained by allowing the lumpectomy cavity to fill
in with serum and fibrin. Drainage of the lumpectomy cavity should be
avoided. Finally, the incision should be closed with a subcuticular
suture to avoid cross-hatching of the skin.
A critical step in lumpectomy is the evaluation of the completeness
of excision of the tumor. To allow adequate histologic evaluation, the
specimen should be removed as a single piece of tissue and should not
be transected unless the pathologist is present. The use of marking
sutures to orient the specimen for the pathologist allows reporting of
the status of individual margins. Gross inspection of the specimen in
the operating room allows identification of positive or close margins,
facilitating immediate reexcision. Frozen-section histologic study is
sometimes useful to evaluate grossly suspicious areas, but the routine
use of frozen sections to evaluate grossly normal margins is of
doubtful value. The ideal amount of grossly normal breast tissue around
the tumor that should be resected as part of a lumpectomy is uncertain.
A resection of 0.5 to 1.0 cm of grossly normal breast tissue results in
histologically negative margins in a large percentage of patients.
[ref: 136] Larger resections may be necessary for invasive ductal
carcinomas with an extensive intraductal component[ref: 309] and for
infiltrating lobular carcinomas. [ref: 310]
When axillary dissection is performed as part of breast-conserving
surgery, a separate incision should be used, except in patients with
tumors high in the tail of the breast. A curvilinear incision at the
edge of the hair-bearing axillary skin provides the best cosmetic
result. The incision should not extend anterior to the fold of the
pectoralis major or posterior to the latissimus dorsi.
The primary indications for a reexcision are positive or unknown
histologic margins of resection on the initial excision. Several
studies have demonstrated residual carcinoma in approximately one-half
of cases when reexcision is performed for positive or unknown
margins.[ref: 308,311] No consensus exists on the best technique for
reexcision. When reexcision is done within 1 to 2 weeks of the biopsy,
it is not usually possible to reexcise an entire biopsy cavity as a
single specimen without sacrificing large amounts of breast tissue. One
author's (M. M.) technique of reexcision in most cases is to reexcise
each of the walls of the biopsy cavity separately. If the initial
specimen is marked with orienting sutures, reexcision can be limited to
the involved margins. Otherwise, thin pieces of tissue are shaved off
each wall of the biopsy cavity and sent as separate specimens, with the
new margin surface marked for the pathologist. When longer intervals
have elapsed between the biopsy and the time of reexcision, contraction
of the biopsy cavity may allow excision of the entire cavity as a
single specimen without sacrificing excessive amounts of breast tissue.
The status of the final margin should be used to determine the
patient's suitability for BCT. Kearney and Morrow [ref: 136] found that
86 of 90 patients undergoing reexcision for positive or unknown margins
were satisfactory candidates for BCT.
There are relatively few complications of breast-conserving surgery.
Wound infection is infrequent, although rates of infection may be
increased when reexcision is performed. The late occurrence of breast
abscess after BCT has been reported. [ref: 312] The median time to
abscess development was 5 months (range, 1.5 to 8.0 months). The only
factor found to correlate with abscess formation was larger size of the
lumpectomy specimen. Cellulitis of the breast occurring at a median of
4 months after BCT also has been reported in approximately 3% of cases.
[ref: 313]

Technique and Complications of Mastectomy

Modified radical mastectomy involves the complete removal of the breast
tissue, the underlying fascia of the pectoralis major muscle, and the
removal of some of the axillary lymph nodes. Modified radical
mastectomy is performed through an elliptical transverse incision,
which encompasses the nipple-areola complex and the biopsy scar if an
open biopsy has been performed. The nipple-areola complex and the
biopsy incision must be removed, but the remainder of the skin of the
breast can be preserved in early-stage breast cancer if needed for
breast reconstruction. With a skin-sparing procedure, additional
exposure to allow complete excision of the breast tissue is achieved by
incision rather than excision of the skin. Skin flaps are created in
the plane between the subcutaneous fat and the underlying breast
tissue. Because of the variability in the amount of subcutaneous fat,
no single thickness is appropriate for all skin flaps. To encompass all
breast tissue, the dissection should extend superiorly to the inferior
border of the clavicle, medially to the lateral border of the sternum,
inferiorly to the superior extent of the rectus sheath, and laterally
to the latissimus dorsi muscle. The fascia of the pectoralis major
muscle can safely be preserved when needed for breast reconstruction.
In general, however, excision posterior to the fascia provides a
convenient plane for ensuring removal of most of the breast tissue.
With the breast attached inferiorly and laterally, axillary dissection
is carried out. Closed suction drains are then placed in the apex of
the axilla and beneath the inferior skin flap. Skin closure is
accomplished with a subcuticular suture. Pressure dressings are not
needed with suction drains and may compromise blood supply.
The term total mastectomy refers to the removal of the entire breast
including the axillary tail of Spence, with preservation of both
pectoral muscles and the axillary nodes. The indications for total
mastectomy include (1) patients with DCIS who elect mastectomy, (2)
patients undergoing prophylactic surgery to prevent the development of
breast cancer, (3) patients who develop a recurrence in the breast
after BCT that had included an axillary dissection, and, on rare
occasion, (4) patients with metastatic disease undergoing mastectomy
for local control of the primary tumor.
Mastectomy is an extremely safe operative procedure. [ref: 314] For
patients with serious comorbid conditions who are at an unacceptably
high risk for general anesthesia, mastectomy can be done using a
combination of local anesthesia, intercostal blocks, and intravenous
sedation. The reported incidence rate of wound infection with
mastectomy ranges from 6% to 14%. [ref: 315,316] The most common
organisms are streptococcus or Staphylococcus aureus. [ref: 316]
Factors that predispose to infection include the use of a two-step
procedure (i.e., open biopsy preceding mastectomy) and prolonged
suction catheter drainage.[ref: 316,317] A single dose of antibiotics
(86% cephalosporin) has been shown to reduce the incidence of wound
infection by 38%. [ref: 318,319] However, because the overall incidence
of infection in these patients is so low, the cost effectiveness of
routine antibiotic prophylaxis for all patients undergoing mastectomy
has not been established. A selective policy of antibiotic
administration to high-risk patients (e.g., prior biopsy, anticipation
of long operating time) seems to be most appropriate. Necrosis of the
skin flaps is a relatively uncommon problem today, but was reported in
8% to 60% of cases, particularly in earlier series of patients
undergoing radical mastectomy. [ref: 320,321] Factors associated with
skin-flap necrosis include denuding the subcutaneous fat from the
flaps, closure under tension, infection, and use of a vertical incision
and occlusive pressure dressings. [ref: 315,316,322] Skin incisions
should be planned to allow tension-free closure. Pressure dressings are
not necessary when suction drains are used, and suspected cellulitis
should be promptly treated with antibiotics. Seroma formation occurs in
100% of patients after mastectomy and should be considered a side
effect, rather than a complication, of the procedure. Prolonged
lymphatic drainage after mastectomy is primarily related to extensive
node involvement followed by obesity and the performance of a two-step
procedure. [ref: 323] Prolonged seroma formation, in addition to
requiring multiple physician visits, may be associated with delayed
wound healing and an increased risk of infection.[ref: 315,316,320]
Seroma formation can be minimized by leaving drains in place until
their combined output is less than 40 mL/24 hours. Fluid collections
should be managed by aspiration every other day. Some advocated the use
of tacking sutures to obliterate the axillary dead space and attach the
skin flaps to the pectoral muscle, [ref: 321] but this has not been
established. Phantom breast syndrome has been long recognized.[ref:
324] In one prospective study,[ref: 325] phantom pain was reported in
13% of patients at 3 weeks, and nonpainful phantom sensations were
present in an additional 15%. Similar incidences were seen at 1 year
and at 6 years. The cause of this syndrome is unknown.

Oncologic Considerations in Immediate Reconstruction

The switch from radical mastectomy to modified radical mastectomy and
advances in plastic surgical technique have made immediate breast
reconstruction an option for most patients who undergo mastectomy. A
number of concerns about immediate reconstruction have been raised and
the available information addressing these concerns comes from
retrospective studies. The incidence of local failure in patients
undergoing breast reconstruction appears to be comparable with patients
treated by mastectomy alone. [ref: 326-328] Similarly, the ability to
detect local recurrence does not appear altered by immediate
reconstruction.[ref: 329-331] There is, however, a greater risk of
problems in patients undergoing immediate reconstruction who require
postmastectomy RT, specifically among patients receiving implants.
[ref: 332,333] This includes patients in whom the implants are placed
beneath latissimus dorsi or transverse rectus abdominis myocutaneous
(TRAM) flaps. In contrast, patients reconstructed with the TRAM flap
without an implant appear to tolerate postoperative RT better. [ref:
334,335] Immediate reconstruction does create technical problems in the
administration of RT. With immediate reconstruction, it can be more
difficult to treat the internal mammary nodes and to boost the chest
wall. Communication before mastectomy between the surgeon and the
radiation oncologist is useful to make decisions on an individual
basis. In summary, immediate breast reconstruction has not been shown
to increase the incidence of local failure or impede the detection of
local recurrence. In the hands of an experienced reconstructive
surgeon, the incidence of complications associated with the procedure
is low, and the need for postoperative systemic therapy should not be
considered a contraindication to immediate reconstruction. In patients
with larger tumors or clinically positive nodes, in whom there is a
high likelihood that postoperative chest wall RT will be administered,
it may be prudent to avoid implant reconstruction or forego immediate
reconstruction altogether.

Indications for Postoperative Radiation Therapy

Postoperative RT refers to the use of irradiation to the chest wall and
draining lymph node regions as an adjuvant treatment after mastectomy.
Postoperative RT has been clearly shown to reduce the rate of local
regional tumor recurrence (i.e., recurrence on the chest wall or in the
axillary, internal mammary, or supraclavicular lymph nodes) by treating
residual microscopic disease that has spread beyond the margin of
surgical resection. In the absence of postoperative RT, there is a
substantial risk of local recurrence after modified radical (or even
radical) mastectomy, principally related to the presence and extent of
axillary nodal involvement. If axillary nodes are involved, local
recurrence is seen in 10% to 30% of patients, whereas if axillary nodes
are uninvolved, local recurrence is seen in only approximately 5% of
patients. [ref: 336,337] Once a local recurrence is clinically
manifest, it can be effectively controlled in only approximately one-
half of patients. [ref: 338-340] Therefore, postoperative RT can
benefit high-risk patients simply by preventing local recurrence.
Despite the clear-cut improvement in local control with adjuvant RT,
[ref: 341] its effect on survival remains controversial. Assessing the
survival value of postoperative RT requires evaluation within large,
prospective randomized clinical trials. There are six published trials
in which patients were randomized after radical, modified radical, or
total mastectomy to postoperative RT or no further treatment in the
absence of systemic therapy.[ref: 342-351] Some of these are among the
earliest clinical trials performed in medicine. In many of these
trials, RT was given using orthovoltage equipment and in most trials
techniques were used that delivered considerable doses to the heart and
are now considered outmoded. Despite this, the use of postoperative RT
clearly reduced the incidence of local recurrence, but none of these
trials demonstrated a clear-cut improvement in the survival rate. In
addition, some of these trials showed a late increase in cardiac
mortality in patients treated with RT compared with unirradiated
patients.[ref: 342,343] The most modern of these trials was conducted
at the Radiumhemmet in Stockholm between 1971 and 1976. [ref: 345,346]
In this trial, 644 patients with operable breast cancer were treated
with modified radical mastectomy and randomized to postoperative RT or
no further treatment. With a median follow-up time of 16 years, node-
negative patients had a decreased rate of local recurrence with
postoperative RT, but there was no effect on distant metastases or
survival. For node-positive patients, the use of postoperative RT was
associated with not only a decrease in local recurrence, but also a
decrease in distant metastasis (P = .02). An overview of randomized
trials of postoperative RT after mastectomy with or without axillary
dissection showed no difference in survival when patients treated with
RT were compared with those treated without RT over the first 10 years
after surgery. [ref: 352] After 10 years, however, there was a lower
rate of survival associated with the use of RT, but this was not
statistically significant. When cause-specific mortality data were
examined, there was an excess of cardiac deaths among patients treated
with RT, but this was offset by a reduced number of deaths from breast
cancer, especially in the more recent trials. An overview reanalysis
published in 2000 showed similar results.[ref: 353] These studies
suggest that if increased cardiovascular mortality associated with
adjuvant irradiation can be avoided by the use of appropriate
techniques, a benefit in survival will be seen.
There are a number of studies that have examined the rate of local
recurrence in patients treated with mastectomy and adjuvant
chemotherapy, but without RT. [ref: 354-357] In the largest of these,
the rate of local recurrence in 2016 node-positive patients entered
into Eastern Cooperative Oncology Group adjuvant systemic therapy
trials was examined. [ref: 356] The 10-year rate of local regional
recurrence (with or without simultaneous distant failure) was 12.9% for
patients with one to three positive nodes and 28.7% with greater than
or equal to four positive nodes. Similar results are observed in the
other studies. [ref: 354,355,357] These studies, as well as the trials
described here, demonstrate a moderate risk of local regional
recurrence in node-positive patients treated with mastectomy and
adjuvant systemic therapy, particularly when four or more nodes are
involved.
A number of studies have examined the issue of adding postoperative
RT to adjuvant chemotherapy (Table 37.2_14). [ref: 357-367] The two
largest of these are from the Danish Breast Cancer Cooperative Group
(DBCG). In DBCG trial 82b, 1708 premenopausal patients who had
undergone mastectomy for pathologic stage II or III breast cancer were
randomly assigned to eight cycles of CMF plus local regional RT or to
nine cycles of CMF alone.[ref: 357] With a median follow-up of 114
months, the 10-year rate of local regional recurrence was reduced from
32% to 9% with RT, and overall survival was improved from 45% to 54%
with RT (both P values <.01) (Fig. 37.2_3). In DBCG trial 82c, 1375
postmenopausal patients who had undergone mastectomy for pathologic
stage II or III breast cancer were randomly assigned to tamoxifen for 1
year plus local regional RT or to tamoxifen alone. [ref: 358] With a
median follow-up time of 123 months, the 10-year rate of local regional
recurrence was reduced from 35% to 8% with RT, and overall survival was
improved from 36% to 45% with RT (both P values <.05) (Fig. 37.2_4).
In a smaller trial from British Columbia, 318 node-positive
premenopausal patients treated with modified radical mastectomy were
similarly randomized to adjuvant CMF chemotherapy and postoperative RT
or to chemotherapy alone. The results of the British Columbia trial
were similar to those of the DBCG trial 82b. [ref: 359] Of note, the
magnitude of the improvement seen in these trials is similar to that
seen with adjuvant systemic therapy (chemotherapy, hormonal therapy, or
both) and suggests that all node-positive patients should receive
postmastectomy RT. A metaanalysis published in 2000 showed a survival
benefit for local regional RT in node-positive women treated with
modified radical mastectomy and adjuvant systemic therapy. [ref: 368]
For a number of reasons, however, these trials have not resulted in the
universal use of postoperative RT in node-positive patients,
particularly those with only one to three positive nodes. One reason is
that the worldwide overview of all trials of postoperative RT only
shows a small, but not statistically significant survival benefit.
[ref: 344] A more substantial reason relates to the issue of
generalizing from these results due to differing surgical and systemic
treatments. The extent of axillary surgery performed in these trials
was less than that performed in the United States, and the rates of
local regional recurrence, especially axillary recurrences, observed in
the Danish trials were greater than observed in U.S. series. In
particular, as noted previously, series of patients with one to three
positive nodes treated by modified radical mastectomy and adjuvant
chemotherapy from the United States show rates of local regional
recurrence in the range of 10%, [ref: 352-356] compared with
approximately 30% in the two Danish trials. Also, systemic therapy in
the Danish trials may have been suboptimal by current standards. What
seems clear is that these trials address an important principle about
the value of establishing local regional control in the presence of
systemic therapy. Without detracting from the importance or validity of
the principle, it is legitimate to question the clinical implications
of these results for practice in the United States. The final reason
for the failure of these trials to translate directly into clinical
practice is the concern about long-term complications of postoperative
RT, especially late cardiac mortality. This is especially relevant in
patients receiving chemotherapy with potential cardiac toxicity, such
as doxorubicin (Adriamycin). To address the issues raised by the
available data on postoperative RT, the American Society of Therapeutic
Radiology and Oncology sponsored a symposium on postoperative RT and
invited a panel to hear the latest information on this topic and to
develop a Consensus Summary Statement. The panel was composed of three
radiation oncologists, a medical oncologist, a surgical oncologist, and
a consumer activist.
The panel agreed to the statements outlined in the next sections.
[ref: 369]

Patient Selection

Postoperative RT may be given to improve local control or to improve
survival. Patients with four or more positive lymph nodes should
receive postoperative RT to improve local control. There may also be a
survival benefit in these patients. The data regarding patient
selection for survival advantage are less clear, but the most recent
evidence suggests that the greatest survival benefit is seen in node-
positive patients with low tumor burdens (i.e., fewer positive nodes or
smaller tumors). Radiation therapy in these patients for survival
benefit is worthy of consideration, pending more definitive data. To
establish the survival benefit of postoperative RT in patients with one
to three positive nodes, a large randomized clinical trial should be
performed. (Such a large Intergroup trial was subsequently approved.)

Technique

Although both recent trials demonstrating a survival benefit with
postoperative RT included treatment to the chest wall, axilla,
supraclavicular area, and the internal mammary nodes, there is
controversy about what sites require treatment. In all patients, the
chest wall should be treated. The value of including the internal
mammary nodes is uncertain and is currently being studied in a large
European randomized controlled trial. In patients with positive
axillary nodes, the internal mammary nodes are known to be also
involved in approximately 30% of cases. Treatment of this area is,
therefore, worthy of serious consideration, provided that it can be
done with acceptable morbidity. Following a level I and II axillary
dissection, the use of a third field to treat the axillary apex and
supraclavicular area is appropriate for selected node-positive
patients, particularly those with four or more positive nodes. A
posterior axillary radiation field is not routinely indicated after a
level I and II axillary dissection. However, if there is concern about
the completeness of surgery, the addition of a posterior axillary field
may be appropriate. Careful attention must be paid to the morbidity of
treatment. In particular, postoperative RT has the potential to cause
late cardiac mortality in patients with left-sided cancers. Therefore,
the amount of heart (and also lung) in the treatment field must be
minimized and documented. Computed tomographic (CT) treatment planning
is useful in accomplishing this goal.

Complications of Radiation Therapy

Possible complications of RT include arm edema, brachial plexopathy,
decreased arm mobility, soft tissue necrosis, rib fractures, radiation
pneumonitis, carcinogenesis (e.g., contralateral breast cancer and
sarcoma), and radiation-related heart disease. Information on heart
toxicity and carcinogenesis is reviewed here. The data on complications
after both postmastectomy RT and RT administered as part of BCT are
reviewed.
Injury to the heart is a possible complication after RT. The long-
term results from the early trials on postoperative RT using outmoded
techniques show an increased risk of cardiac mortality. A more recent
population study showed an increased risk of fatal myocardial
infarctions for patients receiving left-sided breast RT compared with
right-sided breast RT [ref: 370]; however, the techniques of RT are not
described. A large institution study on patients treated with modern
techniques (without adjuvant chemotherapy) did not show an increased
risk with long follow-up. [ref: 371] In the Stockholm postmastectomy
trial discussed previously, late cardiac mortality was seen, but this
was clearly related to the amount of heart irradiated. [ref: 372] Late
cardiac mortality has not been seen in institution-based studies using
limited cardiac irradiation with CMF chemotherapy, [ref: 373,374] but
may possibly occur with RT and high-dose doxorubicin. [ref: 375] CT
simulation allows for three-dimensional visualization of anatomic
structures including the heart and can be useful in selecting
techniques to greatly limit the volume of irradiated heart. [ref: 376]
One possible complication of RT for breast cancer is the induction of
another breast cancer. Breast tissue is known to be sensitive to
radiation carcinogenesis. The latency period between exposure and the
detection of induced cancers is at least 5 years, and this risk
persists for many decades. The risk of carcinogenesis increases with
doses up to 10 Gy, then seems to level off and decline so that for
doses in the therapeutic range (greater than 45 Gy), the risk seems to
be small.[ref: 377,378] Because the dose to the opposite breast from a
course of RT is in the range of 1 to 3 Gy, tumor induction in the
contralateral breast is of greater potential concern than in the
ipsilateral breast. Age at exposure to radiation is the other important
risk factor for carcinogenesis in human breast tissue.[ref: 378-380]
The highest risk occurs in female subjects exposed at the youngest age.
With increasing age, the risk of carcinogenesis declines. The risk for
women older than 40 years of age seems to be negligible, but not zero.
A number of studies have addressed the risk of contralateral breast
cancer after postoperative RT. Boice and colleagues conducted a case-
control study in a cohort of 41,109 patients diagnosed with breast
cancer between 1935 and 1982 in Connecticut. [ref: 381] Among women who
survived for at least 10 years, RT was associated with a small, but
marginally significant, elevation in the risk of a contralateral breast
cancer (relative risk, 1.33). The increased risk associated with RT was
evident among women who were younger than 45 years of age when they
were treated (relative risk, 1.59), but not among older women (relative
risk, 1.01). In another case-control study from Denmark,[ref: 382] the
incidence of a second primary breast cancer in the contralateral breast
was examined among 56,540 women with a first primary breast cancer
diagnosed between 1943 and 1978. In that study, RT was not associated
with an increased risk of contralateral breast cancer (relative risk,
1.04). It is possible to estimate the increased absolute risk of
contralateral breast cancer given the elevated relative risk for
patients aged 45 or younger seen in the study by Boice et al.[ref: 381]
The risk may actually be lower, as is suggested by the Danish study.
For patients aged 45 years or younger, the risk of contralateral breast
cancer within 15 years is approximately 11% without RT and increased to
approximately 12% to 13% with the addition of RT. The risk of
contralateral breast cancer within 30 years is approximately 22%
without RT and increases to approximately 25% to 26% with RT. Thus, the
risk of contralateral breast cancer in young patients, which is already
higher than that for older patients, is likely to be further increased
to a small extent by the use of RT.
Sarcomas of soft tissue or bone are a rare, but well-documented
complication of RT. In a report from Memorial Sloan-Kettering Cancer
Center, a total of 48 patients with a prior history of breast cancer
and subsequent treatment-related sarcomas were seen at the institution
over a 43-year period.[ref: 383] Lymphangiosarcoma of the extremity
accounted for 22, or 46%, of the series. Most of these patients had
been treated with radical mastectomy and postoperative RT. The other 26
sarcomas occurred within a RT treatment field and were considered
radiation related. Twenty-one of these 26 patients were diagnosed with
a soft tissue sarcoma and 5 with a bone sarcoma. The median latency
interval between the diagnosis of breast cancer and the development of
sarcoma was 11 years (range, 4 to 44) and was similar for the two types
of sarcomas. The survival of these patients was poor. In a registry
study from Sweden, 13,490 patients with breast cancer diagnosed between
1960 and 1980 were followed through 1988.[ref: 384] Nineteen sarcomas
were reported, whereas 8.7 were expected (relative risk, 2.2; absolute
risk, 1.7/10**4 person-years). Twelve of the sarcomas appeared within
the radiation fields. Of particular interest is that a high percentage
of sarcomas after treatment with CS and RT are angiosarcomas. [ref:
385,386] The typical initial clinical appearance of this sarcoma is as
reddish, purplish, or bluish nodules or as skin discoloration. There
are often many such areas. These initial findings can be subtle, and
diagnosis is commonly not made until more advanced signs are present.
Of note, the latency period for this radiation-induced sarcoma is
shorter than for other sarcomas, with some occurring before 5 years.
Lung cancer also appears to be increased in patients irradiated for
breast cancer. The latency period from RT to diagnosis is approximately
10 years. In a case-referent study reported by Inskip and others [ref:
387] using the Connecticut Tumor Registry, there was a small increased
risk of lung cancer among 10-year survivors of breast cancer associated
with the use of postoperative RT (relative risk, 1.8; 95% confidence
interval, 0.8 to 3.8). The relative risk for the period 15 years or
more after RT was 2.8 (95% confidence interval, 1.0 to 8.2); however,
it can be difficult to distinguish a late metastasis of breast cancer
to lung from a new primary lung cancer even with a tissue specimen.
This increased relative risk suggests that there would be approximately
nine cases of RT-induced lung cancer per year among 10,000 breast
cancer patients who survived at least 10 years. Similar findings were
noted in a study using the SEER database. [ref: 388] Another study from
this same group suggested that the effects of smoking and RT are
multiplicative,[ref: 389] although this relation has not been
established with certainty. [ref: 390] These studies all examined the
risk of lung cancer after local regional irradiation, and the risk may
be less after more localized breast irradiation.

Sequencing of Systemic Therapy and Radiation Therapy

Clinicians are now commonly faced with the necessity of combining
systemic therapy and RT in patients after CS and after mastectomy.
Given (1) the variable agents, duration, and intensity of CT regimens,
(2) variations in RT and surgical techniques, and, chiefly, (3) the
limited data from randomized clinical trials addressing this question,
the optimal sequencing approach is unresolved. The major goal in
sequencing is to obtain the highest rate of survival; however,
additional important goals are to maintain a low rate of local
recurrence and a low rate of complications. The options for combining
RT and CT are (1) CT first followed by RT; (2) RT first followed by CT;
(3) RT and CT simultaneously; or (4) some number of cycles of CT, then
RT, and then more CT (commonly referred to as sandwich therapy). In
considering this issue, it would be useful to know whether a delay in
either CT or RT decreases its effect, what the complication rate is for
each sequence option, and whether prior RT affects the ability to give
maximal doses of CT. It seems plausible that delays in the initiation
of CT will decrease its effectiveness; however, firm data demonstrating
this are not available. As previously described, the use of
preoperative CT has not clearly improved survival compared with
conventionally timed adjuvant chemotherapy. [ref: 391] Retrospective
reviews of patients treated either with mastectomy or BCT examining the
influence of the delay of CT on outcome have demonstrated conflicting
results.[ref: 392-394] The information available on whether a delay in
the initiation of RT to give CT first leads to a increased local
recurrence rate is also conflicting. [ref: 394-401] The results from
two NSABP adjuvant studies in node-positive breast cancer (B-15 and B-
16) did not show any effect with delays of 12 weeks. [ref: 402]
Prospective randomized clinical trials are required to test formally
the effect of sequencing of CT and RT on outcome. The Dana-Farber/Beth
Israel/JCRT trial is the only published trial specifically designed to
address this issue. In this trial, 244 patients at moderate or high
risk for relapse (almost all node-positive) were randomly assigned to
receive either (1) RT followed by cyclophosphamide, doxorubicin,
methotrexate (with leucovorin), fluorouracil, and prednisone given
every 21 days for four cycles or (2) CT followed by RT. [ref: 403] With
a median follow-up of 58 months, the 5-year crude incidence of first
sites of recurrence suggested that local recurrence was greater with
delayed RT and distant recurrence was greater with delayed CT. The
overall actuarial rate of distant failure was higher in the RT-first
arm (37% vs. 25%; P = .05), thus favoring the CT-first arm. This trial
does not specifically address the question of sequencing in node-
negative patients or with longer than 12 weeks of chemotherapy. One can
hypothesize that the effect of delay in the initiation of RT is related
to the extent of the breast surgery and the final margin status;
namely, a delay of 12 to 16 weeks is important only for patients with
limited breast surgery and close or positive margins. This hypothesis
is suggested by a subset analysis of the JCRT trial; however, this
trial was not large enough to allow for such analysis. At this writing,
there are two ongoing unpublished trials in France. In the Arcosein
trial, approximately 700 patients treated with breast-conserving
surgery will be randomized either to (1) mitoxantrone,
cyclophosphamide, and fluorouracil given every 3 weeks for six cycles
followed by RT, or (2) concurrent CT and RT using the same drugs. In
the other trial, patients are randomized either to (1) epirubicin,
cyclophosphamide, and fluorouracil followed by RT, or (2) concurrent
mitoxantrone, cyclophosphamide, and fluorouracil and RT.
Another important question regarding sequencing is whether RT and CT
can be given simultaneously without an increase in complications or a
decrease in the cosmetic outcome. The use of simultaneous RT and CT has
the advantage of eliminating the necessity for delaying one of the
modalities and perhaps providing an additive or synergistic interaction
between the RT and CT. It has been reported, however, that the
concurrent use of CMF chemotherapy and full-dose RT can result in
greater skin reactions compared with patients treated
sequentially,[ref: 404,405] as well as a decrease in the long-term
cosmetic result. [ref: 404] It may be possible to combine CMF and RT in
other, more tolerable ways. At the University of Pennsylvania, 210
patients were treated with concurrent CF and RT, followed by six cycles
of CMF, with excellent results. [ref: 406] The JCRT has conducted a
prospective pilot study of a modified concurrent CMF and RT regimen
using reduced doses of RT designed to lessen treatment side effects by
anticipating the known interaction of CMF and RT.[ref: 407] One hundred
twelve patients with zero to three positive lymph nodes were entered
into this prospective study. Patients received six cycles of CMF given
every 28 days. On day 14 of cycle 1, patients started tangential field
RT, consisting of 39.6 Gy in 22 fractions to the whole breast and a 16-
Gy boost to the tumor bed using electrons. The most common acute
toxicity observed during or shortly after RT was moist desquamation
(seen in 50% of patients). Grade 4 neutropenia was noted in 16 patients
during RT, but only one patient required hospitalization. Radiation
pneumonitis (grade 2) was noted in only one patient. Fifty-one patients
were evaluable for cosmetic scoring by JCRT physicians 2 years (+/- 6
months) after the end of CT: 47% had excellent, 43% had good, and 10%
had fair cosmetic scores. Seventy-nine percent of patients (89 of 112)
were evaluable for CT dosages delivered; overall, 93% of patients
received at least 85% of all drug doses. Thus, it seems feasible to
consider the administration of concurrent treatment with CMF and
reduced doses of RT; however, longer term data are required to
substantiate this.
Doxorubicin and cyclophosphamide (AC) given every 3 weeks for four
cycles are currently used more widely than CMF, given its shorter
course and at least equivalent outcome. With the substantial
interaction between doxorubicin and RT, it does not seem feasible to
combine these modalities concurrently even with reduced doses of RT.
When AC is used, the results of the JCRT Upfront-Outback trial provide
strong support for the use of AC x 4 followed by RT in all moderate-
and high-risk patients. Preliminary results suggest that AC x 4
followed by Taxol x 4 may further improve outcome compared with AC x 4
alone; the optimal timing of RT in this setting is uncertain. Given
that the positive results in this trial were obtained using RT after
completion of Taxol, it generally seems best to use this approach.
[ref: 408] Studies are under way testing the feasibility of AC x 4
followed by concurrent Taxol and RT; however, there is a report of
early-onset pneumonitis using this approach [ref: 503] and another
report showing none. [ref: 409]

Axillary Treatment

Axillary treatment in the form of a complete dissection was, for many
years, standard management in patients with invasive breast cancer. As
dictated by the Halstedian concept of breast cancer spread, axillary
dissection was considered a critical component of the surgical cure of
the disease. The axillary nodes were considered the filter before
spread of cancer cells to distant sites. Axillary dissection is also
known to be useful in assessing prognosis and ensuring local tumor
control in the axilla. By the 1970s, there was increasing evidence that
axillary dissection had a limited effect on survival. This was most
convincingly demonstrated in the NSABP trial B-04. In this trial,
patients with clinically negative axillary nodes were randomized to
radical mastectomy, total mastectomy with observation of the axillary
nodes and a delayed dissection if positive nodes appeared, or total
mastectomy with RT to the regional lymphatics. [ref: 234] No
statistically significant difference in the 10-year survival rate was
found among the groups, despite the fact that approximately 40% of the
patients undergoing axillary dissection had positive nodes and a
similar percentage were presumed to have positive nodes in the
observation-only arm. While this study emphasizes that axillary
dissection is not of survival benefit for the majority of patients, the
number of patients in this trial was insufficient to rule out a small,
but clinically important survival benefit. [ref: 410] Also of note is
that 25% to 30% of the 20-year survivors reported in a series of
patients treated with radical mastectomy alone had positive axillary
nodes, [ref: 411-414] suggesting that, for a small number of patients,
axillary dissection may be therapeutic. The therapeutic potential of
axillary dissection is also supported by the survival benefits noted in
the more recent trials of postmastectomy RT. [ref: 415-417]
With the general recognition that axillary dissection was principally
a prognostic, rather than therapeutic, procedure, a number of studies
were undertaken to determine the extent of axillary surgery needed to
determine whether nodes were positive or negative. Many of these
studies examined the likelihood of skip metastases (i.e., involvement
of nodes in the upper axilla; level III) in the absence of involvement
in the lower (level I or II) nodes. Involvement of level III is clearly
rare when both levels I and II are negative. [ref: 418,419] There is
considerable variability in the literature on the risk of skip
metastases to level II. Much of this variability may be due to
variations in the definition of which nodal tissues constitute levels I
and II. These disparate observations have led some authors to conclude
that a level I dissection provides accurate staging information, [ref:
420] while most have concluded that removal of both levels I and II is
required. [ref: 421,422] A level I and II dissection is effective at
providing local control in the axilla. Among patients treated with a
level I and II dissection as part of BCT, axillary recurrence rates of
less than 3% have been reported.[ref: 423-426] When patients undergo
more limited random axillary sampling procedures, the likelihood of
local recurrence is related to the number of lymph nodes removed. The
5-year probability of an axillary recurrence is approximately 20% in
patients with no lymph nodes examined and approximately 10% when only
one to two negative nodes are removed. At least six to ten nodes need
to be removed to avoid misclassification and to optimize local control
in the axilla. [ref: 234,426-429] While a level I and II dissection is
generally well tolerated, there are occasionally complications. Major
complications, including injury or thrombosis of the axillary vein and
injury to the motor nerves of the axilla, are infrequent. Minor
complications are more common and include seroma formation, shoulder
dysfunction, loss of sensation in the distribution of the
intercostobrachial nerve, and edema of the arm and breast.
A number of developments have led to a reexamination of the need for
axillary dissection in all patients. Under discussion is the routine
use of adjuvant systemic therapy in many patients with node-negative
breast cancer, the increasing use of BCT, and the increasing number of
patients with small, mammographically detected cancers with a low risk
of axillary metastases. One approach to avoiding axillary dissection is
to identify cancers with a low risk of nodal metastases. The incidence
of axillary nodal involvement is known to be related principally to
tumor size. However, axillary node metastases are still seen in 12% to
37% of cancers measuring 1 cm or less, [ref: 170,429-435] and in a
number of studies, the incidence of metastases does not decrease
appreciably even with cancers 0.5 cm or smaller. [ref: 429,432,434] The
only groups of patients with invasive carcinoma regularly identified as
having nodal metastases in fewer than 5% of cases are those with
microinvasive tumors, those with grade 1 tumors less than 5 mm,[ref:
415] and those with pure tubular carcinomas less than 1 cm. [ref:
436,437] However, most patients continue to undergo axillary
dissection, even in these favorable groups. A Patterns of Care study
examining axillary surgery in 17,151 patients with stage I and II
carcinoma undergoing BCT in 1994 found that overall, 93% had an
axillary dissection. Even in patients with grade 1 tumors, favorable
histologic subtypes, or tumors less than 5 cm, 88% or more underwent
axillary dissection.
The new technique of lymphatic mapping and sentinel node biopsy
offers the possibility of reliably identifying patients with axillary
node involvement with a low-morbidity operation, allowing axillary
dissection to be limited to patients with nodal metastases who can
benefit from the procedure. The sentinel node is defined as the first
node receiving lymphatic drainage from a tumor, and the absence of
metastases in the sentinel node reliably predicts the absence of
metastases in the remaining axillary nodes. This concept was
popularized by Giuliano et al. in malignant melanoma and later adapted
to breast cancer. [ref: 438] Multiple studies have now confirmed that a
sentinel node can be identified in more than 90% of cases, with
experience, and predicts the status of the remaining nodes with 90% to
95% accuracy [ref: 439-444] (Table 37.2_15).
The sentinel node can be identified using isosulfan blue dye,
radiolabeled colloids, or a combination of the two agents. Similar
success rates are reported for the techniques, and a randomized trial
demonstrated no difference in the rate of sentinel node identification,
predictive value of the sentinel node, or learning curve for isosulfan
blue dye alone compared with the blue dye plus technetium sulfur
colloid.[ref: 445] Approximately 20 to 30 cases appear to be necessary
to master the sentinel biopsy technique, with individual learning
curves varying widely. [ref: 444-446]
A number of contraindications to sentinel node biopsy have been
identified, including the presence of suspicious axillary adenopathy,
evidence of locally advanced breast cancer, use of preoperative
chemotherapy, multifocal or multicentric tumor, prior axillary surgery,
and a pregnant or lactating patient. In addition, virtually no
information on accuracy of the procedure for tumors larger than 5 cm is
available. Lymphatic mapping and sentinel node biopsy have prompted
renewed interest in the role of the internal mammary nodes in breast
cancer. A small number of internal mammary node biopsies have been
reported in patients whose lymphoscintigraphy results have failed to
demonstrate axillary drainage, but the need for internal mammary node
biopsy remains to be defined.
Sentinel node biopsy offers the pathologist the opportunity to
perform a much more detailed examination of one or two nodes than is
possible when evaluating an entire axillary specimen. It has been known
for more than 30 years that approximately 20% of lymph nodes in which
no tumor is seen after routine processing and light microscopy contain
tumor cells that can be identified by serial sectioning or
immunohistochemistry. [ref: 447-450] Early studies did not find these
micrometastases to be prognostically significant, and these techniques
were not practical for use on an entire axillary specimen. With the
ability to examine only one or two sentinel nodes, there is renewed
interest in the possibility of using immunohistochemistry for ultra
staging. However, prospective confirmation of the significance of tumor
cells detected by immunohistochemistry is lacking. The retrospective
studies that indicate prognostic significance show a wide variation in
the magnitude of this effect, [ref: 451-453] making patient counseling
difficult. The results of two prospective clinical trials being carried
out by the American College of Surgeons Clinical Oncology Group and the
NSABP to address these issues will provide important information
regarding the clinical significance of micrometastases.
An additional important unresolved issue is the need for completion
axillary dissection when a positive sentinel node is identified. The
initial sentinel node studies demonstrated that the sentinel node is
the only tumor-containing node in 40% to 60% of cases, and fewer than
20% of patients have more than three involved nodes. [ref: 438-445] A
positive sentinel node establishes the need for systemic therapy, and,
in patients undergoing breast-conserving surgery, radiation of the
breast includes a significant portion of the low axilla, helping to
maintain local control. However, the possibility that axillary
dissection has some therapeutic benefit for patients with involved
nodes cannot be excluded, making abandonment of completion axillary
dissection unwise. The need for axillary dissection in sentinel node-
positive patients is being addressed in a prospective randomized trial
by the American College of Surgeons Clinical Oncology Group.
The initial results of lymphatic mapping and sentinel node biopsy are
extremely promising. However, data on the long-term outcome of sentinel
node biopsy alone in unselected populations and information on the
ability of surgeons outside of centers of expertise doing a low volume
of breast surgery are needed before it can be determined if sentinel
node biopsy will replace axillary dissection as the standard of care
for the node-negative or node-positive breast cancer.