rad ther tech

Radiation Therapy Techniques for the Intact Breast

After wide local excision, segmental mastectomy, or quadrantectomy,
the breast is irradiated with lateral and medial tangential portals.

Treatment Volume

The entire breast and chest wall with a small portion of underlying
lung should be included in the irradiated volume. Radiopaque surgical
clips placed at the margin of the tumor bed may assist in defining the
target volume. [ref: 710] When it is combined with a supraclavicular
portal, the upper margin of the portal is placed at the second
intercostal space (the angle of Louis). If the regional lymph nodes
are not to be irradiated (as in patients with intraductal disease or
negative axillary lymph nodes or when adjuvant chemotherapy is
administered in some patients with positive axillary lymph nodes), the
upper margin of the portals should be placed at the head of the
clavicle to include the entire breast (Fig. 50-19). The medial margin,
if no internal mammary portal is used, should be 1 cm over the
midline. If an internal mammary field is used, the medial tangential
portal is located at the lateral margin of the internal mammary field
(Fig. 50-20). The lateral-posterior margin should be placed 2 cm
beyond all palpable breast tissue, which is usually near the
midaxillary line. The inferior margin is drawn 2 to 3 cm below the
inframammary fold.

In selected patients, the standard portals must be modified to
accommodate topographic features of the patient or location of the
tumor excision site to avoid junction of fields at the scar
(Fig. 50-21). In some patients, the breast falls superiorly toward the
supraclavicular area in the supine position. An inclined board placed
on the treatment table or a thermoplastic mold can correct this
problem.

In patients with wide tangential fields bridge separation (> 22 cm) to
which treatment is delivered with 6-MV or lower-energy photons, there
is significant dose inhomogeneity in the breast (Fig. 50-22); as noted
by Moody and colleagues [ref: 510] and Taylor and co-workers, [ref:
757] this is correlated with less satisfactory cosmetic results. This
problem can be minimized by using higher-energy photons (10 to 18 MV)
to deliver a portion of the breast irradiation (about 50%), as
determined with prospective treatment planning to maintain the
inhomogeneity throughout the entire breast to 5% or less. If desired,
the buildup of the beam may be modified with a "degrader," or, as done
at Washington University Medical Center, a thermoplastic mold may be
constructed to support the breast in the treatment position
(Fig. 50-23).

Other treatment positions have been used to improve the dosimetry in
patients with large, pendulous breasts. At the Institut Curie, these
women are treated in the lateral decubitus position to flatten the
breast contour. The patient's position allows the breast to flatten
over a support, providing a rather homogeneous thickness throughout
the treated volume. The dose to midthickness on the beam axis can
easily be determined from entrance and exit dose measurements. [ref:
249] A modified lateral decubitus position with an immobilization
device has been suggested (Fig. 50-24A). [ref: 132] The dose
distribution with a "mock wedge" is satisfactory (Fig. 50-24B).
Irradiation in the prone position has been proposed by Merchant and
McCormick, [ref: 494] with reduction of dose in the high-dose region
to 102% to 103% of the dose to the irradiated breast, as well as
reduction of volume and dose to the underlying lung and heart and
reduction of scattered dose to the contralateral breast.

A polyvinylchloride, ring-shaped device, held by a strap, has been
used around the breast to aid in positioning of patients with large
pendulous or flaccid breasts. [ref: 51]

Alignment of the Tangential Beam with the Chest Wall Contour

The anterior chest wall in most women slopes downward from the
midchest to the neck. To make the posterior edge of the tangential
beam follow this downward-sloping contour, the collimator of the
tangential beam is usually rotated. An alternative technique is used
at the Mallinckrodt Institute of Radiology: the posterior edge of the
tangential beam is made to follow the chest wall contour by means of a
rotating beam splitter mounted on a tray without rotation of the
collimator (Fig. 50-25). In this way, the superior edge of the
tangential beam remains in the true vertical and matches perfectly the
vertical inferior edge of the supraclavicular field. Examples of
localization films for tangential portals are shown in Figure 50-26.

Usually 2 to 3 cm of underlying lung is included in the tangential
portals. The amount of lung included in the irradiated volume is
greatly influenced by the portals used. Roberson and colleagues, [ref:
612] using CT data, demonstrated the impact of tangential breast
fields (medial border 1 cm or 3 to 4 cm beyond midline) and internal
mammary portals on the percentage of lung volume irradiated using
photons or electrons (Fig. 50-27).

Bornstein and associates [ref: 72] determined the amount of lung
irradiated in 40 patients with breast cancer using CT scans for
treatment planning in the treatment position. Parameters measured from
simulator films included the perpendicular distance from the posterior
tangential field edge to the posterior part of the anterior chest wall
at the center of the field (central lung distance, CLD), the maximum
perpendicular distance from the posterior tangential field edge to the
posterior part of the anterior chest wall (maximum lung distance,
MLD), and the length of lung (L) as measured at the posterior
tangential field edge on the simulator film (Fig. 50-28). The
ipsilateral total lung area and the lung area included within the
treatment portal were calculated for each CT scan slice, multiplied by
the slice thickness, and integrated over all CT scan slices to give
the volumes. The best predictor of the percentage of ipsilateral lung
volume treated by the tangential fields was the CLD. A CLD of 1.5 cm
predicted that about 6% of the lung would be included in the
tangential field; a CLD of 2.5 cm predicted about 16% and a CLD of 3.5
cm about 26% of the ipsilateral lung (Fig. 50-29).

Chen and associates [ref: 107] developed a model based on simulation

measurement data to predict the lung geometry included in tangential
fields. This model is used for calculating dose distributions with
inhomogeneity corrections, which potentially decrease pulmonary
toxicity.

Special attention should be paid to minimizing the volume of heart
irradiated. Fuller and colleagues [ref: 275] retrospectively collected
data on radiation doses to the heart and coronary arteries during
breast irradiation. CT scans of the chest were obtained, and
three-dimensional (3-D) reconstruction was made of the heart, lung,
and body contour. Original dose distributions were superimposed on
these outlines, and doses to the total cardiac volume and to three
main coronary arteries were estimated using an alpha/beta ratio of 4
Gy. Nine of 14 patients had mastectomy followed mainly by orthovoltage
irradiation with similar techniques used up until 1984. Thirteen of 14
patients treated more recently had conservative surgery followed by a
modern two- or four-field megavoltage technique. For 20 patients with
left-sided tumors, the heart volume irradiated to a minimum
extrapolated target dose of 50 Gy was significantly decreased with
modern techniques.

Doses and Beams

Minimal tumor doses of about 50 Gy are delivered to the entire breast
over 5 to 6 weeks (1.8- to 2-Gy tumor dose daily, five weekly
fractions). Minimal doses of 46.8 Gy (1.8-Gy daily fraction) are
preferred for patients with large, pendulous breasts or when
irradiation is combined with chemotherapy. [ref: 331] At Washington
University Medical Center, the prescribed central axis dose (46.8 to
48.6 Gy) is at the midpoint of the tangential fields (bridge)
separation.

X-ray energies of 4 to 6 MV are preferred to treat the breast. Photon
energies greater than 6 MV may underdose superficial tissues beneath
the skin surface, but higher-energy photons may be helpful in large
breasts to decrease the integral breast dose. Chin and associates,
[ref: 108] in a breast phantom study with various energy photons,
demonstrated that the areas of high dose (hot spots) noted along the
periphery of the breast occurred less frequently with 6- and 8-MV
photons. At the same time, a decreased dose was delivered to the skin
and superficial portions of the breast (Fig. 50-30). Inhomogeneity in
the dose distribution (as much as 25% from central axis dose) is noted
at the inferior portion of the breast, where the entrance-exit
distance of the two tangential beams is significantly shorter than at
the central plane. This is not easily appreciated on a two-dimensional
treatment plan; fortunately no significant fibrosis is observed in
this area.

Wedges or compensating filters must be used for a portion of the
treatment to achieve a uniform dose distribution within the breast (5%
to 8% dose variance from the chest wall to the apex) (Fig. 50-31AB and
Fig. 50-31CD). Evans and associates [ref: 197] described a method to
design compensators for breast irradiation using electronic portal
images to obtain a "radiological thickness map" of the breast. This
mapping is based on pseudo-CT slices modeled to determine the amount
and optimum wedge angle for compensation.

It is not necessary to apply bolus to the breast (except with photon
energy > 6 MV), because the skin is usually not at risk for recurrence
after complete excision of a T1 or T2 lesion as is the skin of the
chest wall after a mastectomy. Use of bolus results in impaired
cosmetic results.

Three-Dimensional Treatment Planning

Solin and colleagues [ref: 708] devised 38 3-D treatment plans in two
patients using multiple CT scan sections and compared various beam
energies and dose distributions. Increased doses (hot spots) were seen
in the central axis plane when lung inhomogeneity corrections were
used, and additional hot spots were noted in off-axis planes toward
the cephalad and caudal portions of the target volume. Breast
inhomogeneity doses ranged from 5% to 10%. **60Co produced greater
inhomogeneities than 6-MV photons with minimal improvement in tumor
dose coverage. In contrast, 15-MV photons had significantly worse
tumor coverage at shallow depths, although there was a slight
reduction in hot spots. These authors were unable to identify any beam
arrangement that improved dose distributions when compared with
tangential fields.

Neal and co-workers [ref: 524] described a 3-D technique in 20 women
treated with breast-conserving surgery and irradiation using CT. A
margin of 5 mm was left between the breast planning target volume and
the skin surface to ensure that the planning target volume did not
fall within the high-dose gradient of the buildup of the photon beam.
With 3-D planning, the mean percentage of ipsilateral lung volume
receiving less than 20% of the isocenter dose was 93.3%, that
receiving 20% to 50% was 1.8%, and that receiving 50% to 80% was 1.3%.
For 11 left-sided breasts planned, the mean percentage of myocardial
volume receiving less than 20% of the dose was 97.6%, that receiving
20% to 50%, 1.1%, and that receiving 50% to 80%, 0.6%. Greater dose
heterogeneity was observed in larger breasts (volume > 500 cc)
(Fig. 50-32). These findings are similar to those with two-dimensional
dosimetry treatment planning.

Another method for 3-D treatment planning in 20 patients treated with
breast conserving surgery was described [ref: 281] using 8 to 12 CT
images per patient and an isocentric technique with tangential fields.

At this time, because of limited 3-D treatment planning resources and
because patients with tumors at other anatomic sites may benefit more
from 3-D treatment planning, we have not considered this approach in
the routine treatment of patients with breast cancer.

Boost to Tumor Site

Fisher and colleagues [ref: 220] questioned the need for a radiation
dose boost at the tumor excision site, yet delivery of a boost dose
increases the probability of local tumor control. Recht and Harris
[ref: 599] discussed the rationale for administration of a boost dose
to a limited volume of the breast that could reduce the incidence of
local recurrence. In 1504 patients, Clark and associates [ref: 116]
observed a 17% incidence of failure at 10 years in patients to whom no
boost was delivered, compared with 11% in patients receiving doses of
5 to 15 Gy at the primary site (P = 0.03). In other series in which
surgical margins were unknown, roughly half the breast failure rate
(6% to 11%) was noted with boost, compared with no boost (9% to 20%).
[ref: 287,603]

The indications for boost irradiation are strongly supported by
pathologic findings described by Holland and colleagues, [ref: 357]
who described the incidence of multifocal carcinoma in breast tumors 2
cm or smaller in diameter as a function of distance from the edge of
the primary tumor: 17% of patients had additional tumor within 1 cm,
28% had carcinoma in situ, and 14% had invasive carcinoma within 2 cm
from the edge of the tumor. Careful pathologic examination of
mastectomy specimens in patients who were candidates for breast
conservation therapy (with margins of 1 to 2 cm) demonstrated that
residual microscopic tumor can be detected within 2 cm of the primary
tumor in 20% to 40% of patients. [ref: 356,357,520] Therefore a
significant residual tumor burden may be present after an excisional
biopsy, even with negative pathologic margins, a finding that
correlates well with the NSABP B-06 breast recurrence rate of 50%
after excision alone with negative margins. [ref: 219]

Most authors report that 65% to 80% of breast recurrences after
conservation surgery and irradiation occur around the primary tumor
site. [ref: 95,242,259,333,410,603] These data provide a strong
rationale for a tumor bed boost, although the issue has not been
resolved. Various series suggest that patients treated with higher
doses have a greater probability of tumor control. [ref: 602,603]

Because of low to nonexistent morbidity with little detriment to
cosmesis, at Washington University a tumor bed boost is administered
to all patients except those with tumors smaller than 1 cm and with
pathologically generous (> 2 cm) margins or quadrantectomy. Doses
ranging from 10 to 20 Gy are administered, depending on size of the
tumor and the status of the excision margins. In the future, certain
subsets of patients may be defined who do not require a boost, such as
women older than 50 years of age who have T1 tumors smaller than 1 cm,
absence of intraductal carcinoma, negative surgical margins, no
necrosis, and low-grade tumors.

Before the widespread availability of electron-beam therapy,
interstitial brachytherapy or cone-down photon boost was popular. Many
institutions currently prefer electron-beam boost because of its
relative ease in setup, outpatient setting, lower cost, decreased time
demands on the physician, and excellent results when compared with
**192Ir implants. Radiation oncologists who prefer the brachytherapy
boost technique point to decreased skin dose and potential
radiobiologic advantages when compared with electron-beam boost
therapy.

Boost with Electron Beam

The patient is positioned with the arm toward the head to flatten the
breast contour, and she is rolled so that the tylectomy scar is
parallel to the table and the accelerator head can point straight down
onto the target volume. An electron energy is selected that covers the
target volume depth (usually 12 or 16 MeV), based on review of the
mammogram to ascertain the location and depth of the tumor or metallic
surgical clips. The 90% prescription isodose line is limited to the
chest wall to decrease dose to the lung. The clinical setup for
electron boost involves marking the projection of the postoperative
scar on the skin and adding 3 cm in all directions (Fig. 50-33A and
Fig. 50-33BC).

Accurate target volume definition is critical with any boost
technique. Methods vary from simple and unsophisticated (as described
in the previous paragraph) to complex and expensive, such as
ultrasound and CT definition of the target volume.

The surgical clip method requires the cooperation of the surgical
team. Despite the fact that it would theoretically take an infinite
number of clips to define every extension of a typical tylectomy
cavity, in practice six clips suffice (superficial, deep, medial,
lateral, cephalad, and caudal). [ref: 710] In a study reported by
Denham and colleagues, [ref: 155] surgical hemoclips were left in situ
in 27 patients to demarcate the limits of the excision cavity. The
position of these clips varied widely in relation to the patient's
recollection of the position of the original lump, the surgical notes,
and the surgical scar. Incomplete coverage of the excision cavity in
the "coronal" (en face) plane using an electron field could have
occurred in an estimated 10 (42%) of 24 cases had surgical clips not
been left in situ. Depth of the surgical clips below the skin surface
also varied markedly; in 19 (73%) of 26 cases, the clips were observed
to be 3 cm or more below the skin surface, whereas in only 5 (19.2%)
of 26 cases were the clips found to be 2 cm or less deep to the
surface. Had a 9-MeV electron beam been used to treat all of the
patients, a major underdose of the excision cavity would have been
likely in 21 of 26 evaluable cases (81%). Coverage would have improved
to 11 (42%) of the 26 had a 12-MeV beam been used.

In a British study, [ref: 326] 50 patients treated with breast
conservation therapy had excision cavity boundaries marked by surgical
clips. An electron-beam boost field was initially planned to achieve a
2-cm margin. When evaluated by radiographs and the position of the
surgical clips, the clinical field was found to be inadequate in 34
patients (68%). However, if 3-cm margins around this scar had been
used, the median percentage of geographic misses would have been only
18.6%. Also, median distance from the center of the scar to the
deepest clip was 3.8 cm, suggesting that electron-beam energies of 12
MeV or higher can adequately cover most excision sites to volume.
Nevertheless, in 10 patients the deepest clip was at 5 cm or greater
depth.

Fein and associates [ref: 207] described a comparison in patients with
stage I or II breast cancer treated with breast conservation therapy;
surgical clips were placed in the excision cavity in 556 patients, and
no clips were placed in 808 patients. After breast irradiation with
tangential fields (46 to 50 Gy in 4.5 to 5 weeks with 4-MV photons)
and, in some patients, lymph node irradiation to similar doses, the
primary tumor incision site was boosted with either electrons or
external beam (14 to 20 Gy). The actuarial breast recurrence rates at
10 years were 5% in patients with clips and 11% in patients without
clips (P = 0.01). Increased rates of breast recurrence were noted for
patients with clips who had some of the following variables: no
adjuvant treatment, unknown surgical margins, no reexcision,
pathologic negative nodes, and outer location of primary tumor. The
higher incidence of breast relapse may be related to a specific
surgeon who had a breast recurrence rate of 21%, compared with 6% for
the remainder of the surgeons (P = 0.01); the status of the margins
was unknown in 48% of his patients, compared with 10% overall (P =
0.001). Excluding this surgeon's patients from analysis, the isolated
breast recurrence rate for patients with clips was 6%, compared with
5% for the group without clips (P = 0.18). The authors concluded that
failure to ink the surgical specimen and inadequate assessment of
margins cannot be compensated by placement of surgical clips or
treatment planning using CT to delineate the surgical bed. On the
other hand, this study failed to show any benefit from use of surgical
clips at the tumor excision margins to design the boost volume.

Ultrasound can provide the depth of the biopsy cavity, as well as the
other dimensions, for use in designing electron portal borders and
selection of electron energy. Ultrasound was used in 30 patients to
measure breast thickness for determination of the most appropriate
electron-beam energy for the boost. In most patients the depth was 4
cm or less, but in 8 patients (32%) energy higher than 12 MeV should
have been used to adequately cover the depth of the target volume.
[ref: 288]

CT-guided portal design must be done in the treatment position. This
technique gives good definition of depth of the chest wall, but it is
not practical for designing the portal edges.

Cost-effective methods to optimally determine the most accurate way to
deliver the boost should be prospectively evaluated, correlating them
with local treatment failures.

Boost with Interstitial Implant

At Washington University Medical Center, selection for brachytherapy
includes (1) women with large breasts and deep tumors (> 4 cm below
skin), since the integral dose with electrons is high and there can be
exit dose into the lung, and (2) patients with microscopically
positive or unknown margins who are not undergoing reexcision or who
have other poor pathologic risk features, since a higher dose can be
more easily delivered at depth with the implant. Brachytherapy has
also been suggested for patients with EIC. [ref: 158]

Ideally with interstitial brachytherapy, the optimal target volume is
determined in the operating room in consultation with the surgeon
(minimum, 1-cm margin). At some institutions, including ours,
intraoperative implants have been performed to reduce cost and enhance
the accuracy of placement of the catheters for the radioisotope. [ref:
469]

We place afterloading catheters at the time of the tylectomy or
reexcision or the axillary dissection; usually two planes (superficial
and deep) cover the volume in T1 and T2 tumor beds. If the implant is
to be carried out later, after the breast irradiation has been
completed, consultation with the surgeon or, preferably, use of
metallic surgical clips is helpful in determining the target volume.
See Chapter 17 for further technical details of brachytherapy.

Comparison of Electron-Beam and Brachytherapy Boost

The decision to use either electron-beam or brachytherapy boost rests
on convenience, radiation safety, and cost considerations. In patients
who can easily come to the radiation oncology facility for daily
treatment, electrons may be preferred. [ref: 604]

Touboul and colleagues [ref: 773] reported on a nonrandomized study of
329 patients with breast cancers 3 cm or less in diameter who received
breast irradiation (median dose, 45.9 Gy) and were given a boost dose
of approximately 15 Gy with either electron beam (9 or 12 MeV) in 160
patients or interstitial **192Ir implant in 169 patients. The 10-year
local breast-relapse rate was 15.5% with electron-beam boost and 8.1%
with interstitial brachytherapy (P = 0.32). Cosmetic results were
recorded as excellent or good in 82% of patients treated with electron
beam and 61% of patients treated with interstitial brachytherapy (P =
0.001).

The incidence of breast relapse in 839 breasts treated at Washington
University Medical Center with either **192Ir implant or electron beam
boost was equivalent (Fig. 50-34A and Fig. 50-34B). We observed
similar results at various dose levels with either modality in T1
tumors; in T2 tumors treated with brachytherapy, one breast relapse
(4%) occurred in 24 patients treated with 70 Gy, in contrast to 4
(28.6%) of 14 patients receiving lower doses.

Mansfield and associates [ref: 470] updated their results in 654
patients treated with interstitial iridium implant boost and 416
patients treated with electron-beam boost. The 10-year local tumor
control rates were 88% with **192Ir brachytherapy and 82% with
electrons in the patients with T1 tumors treated with either technique
(91% and 93% at 5 years, respectively). In stage II patients, the
tumor control rates were 98% with interstitial implant and 85% with
electron-beam boost (P = 0.02). The 10-year disease-specific survival
rates were 90% for stage I and 69% for stage II tumors. Cosmetic
results were judged as excellent or good in 91% of patients treated
with **192Ir and in 95% of those receiving electron-beam boost.

Vicini and colleagues [ref: 815] reviewed treatment outcome in 402
cases of stage I or II breast cancer treated with breast conservation
surgery and 45 to 50 Gy to the breast followed by a boost to the tumor
bed using either electrons (104 patients), photons (15 patients), an
**192Ir interstitial implant (197 patients), or an **125I implant (86
patients) to at least 60 Gy. With a median follow-up of 59.3 months,
no differences in the 5-year actuarial rate of local recurrence were
noted among patients boosted with any of the four methods (5.4%, 0%,
3.8%, and 3%, respectively). Also, no significant differences in the
percentage of patients obtaining good or excellent cosmetic results
were noted (90%, 82%, 88%, and 93%, respectively).

A few reports have compared outcome in patients treated with either
electron-beam or brachytherapy boost. Fourquet and associates [ref:
248] described a randomized study in which 255 patients with 3- to
7-cm tumors were treated with breast irradiation (50 Gy) and a 20-Gy
boost with reduced tangential **60Co fields or an interstitial iridium
implant. The 8-year cumulative risk of breast recurrence was 39% with
**60Co and 24% with interstitial brachytherapy. Cosmetic evaluation
carried out in 120 patients showed satisfactory cosmesis in 75% of the
**60Co and 71% of the **192Ir group.

Results reported by several authors are summarized in Table 50-24.

The Radiation Therapy Oncology Group (RTOG) initiated a randomized
comparison of photons versus electrons; 295 patients were registered,
but the data have not been published. [ref: 604] The EORTC is
conducting a randomized trial to assess the role of the boost dose in
breast-conserving therapy in stage I and II breast cancer. Details of
the protocol, dosimetry, and quality assurance procedures were
reported, [ref: 796] but results have not been published.

Cosmesis with Electron or Brachytherapy Boost

Pezner [ref: 567] noted that fibrosis compromises cosmetic results in
breast conservation therapy, and it is usually related to the use of a
local boost. Analysis of 449 patients given a boost dose with
electrons and 129 boosted with **192Ir interstitial brachytherapy at
our institution showed that boost type or dose had no impact on
cosmetic results. [ref: 562] In patients with T1 tumors, cosmesis was
excellent or good in 84% of patients receiving electron-beam boost and
81% of patients treated with brachytherapy; with T2 tumors, the
respective figures were 74% and 79%. There was a suggestion that a
total dose to the boost volume of more than 65 Gy (with either method)
yielded less satisfactory cosmetic results. However, the differences
were not statistically significant.

Ray and Fish [ref: 589] reported excellent cosmetic results in 91% of
107 patients given a boost dose with electrons, compared with 52% of
23 patients treated with an interstitial iridium implant. Olivotto and
associates, [ref: 536] in 497 patients receiving a boost by
interstitial implant, reported excellent cosmesis at 3 years in 58%,
compared with 85% of those given a boost with photons or electrons and
85% of those given no boost dose (P = 0.03). This result may be
related to the volume implanted and dose delivered.

Cosmesis results with either boost technique at various institutions
are summarized in Table 50-25. Matching the Tangential Fields with the Supraclavicular Field

A hot spot caused by divergence of the tangential beams into the
supraclavicular field and of the supraclavicular beam into the
tangential fields can exist just beneath the skin surface at the
junction of the inferior border of the supraclavicular field and the
superior border of the tangential fields. [ref: 47] The sharp beam of
a linear accelerator and the "horns" at the edge of this beam produce
a marked increase in dose beneath the matchline if these divergences
are not corrected. This increased dose may result in severe matchline
fibrosis or even rib fracture.

The divergence of the tangential fields can be eliminated by angling
the foot of the treatment couch away from the radiation source to
direct the tangential beams inferiorly so that the superior edges of
these beams line up perfectly with the inferior border of the
supraclavicular field (Fig. 50-35). [ref: 744] In addition, the
collimator must be rotated to geometrically eliminate overlap at this
junction; alternatively, the "hanging block" technique developed at
the Harvard Joint Center, in which a vertical block is affixed to the
superior portion of the collimator to block off the nonvertical
portion of the tangential beam, can be used (Fig. 50-36). The inferior
divergence of the supraclavicular beam can be eliminated by blocking
off the inferior half of this beam with a beam splitter so that the
central, nondiverging portion of the beam becomes the inferior border
of this field (Fig. 50-36).

Matching the Tangential Fields with the Internal Mammary Field

When an internal mammary field is required, the match between it and
the medial tangential field can be a problem if there is a significant
amount of breast tissue beneath the matchline. In this situation, a
cold spot can exist (Fig. 50-37, A). The effect may be negligible if
the breast tissue beneath this matchline is thin (Fig. 50-37, B), or
it can be avoided by not using a separate internal mammary field
(Fig. 50-37, C). In the latter case, the internal mammary nodes must
be included in the tangential beams (as determined by CT scan or
radionuclide scintigraphy). Usually this can be achieved by moving the
medial tangential field border 3 to 5 cm across the midline. The
portal films should be inspected carefully to ensure that an excessive
amount of lung or heart is not being irradiated. There is no good
solution to this matchline problem in large-chested women who also
have a significant amount of breast tissue beneath the matchline of
the tangential and internal mammary fields. Woudstra and van der Werf
[ref: 861] described a technique using an oblique incidence of the
internal mammary portal to match the orientation of the adjacent
medial tangential portal; this results in a more homogeneous dose
distribution at the junction of the two fields (Fig. 50-38).

Irradiation Dose to the Contralateral Breast

Fraass and colleagues [ref: 267] measured the radiation dose to the
contralateral breast in 16 women treated with tangential fields and
performed phantom measurements. For a typical treatment of 50 Gy, the
contralateral breast received 0.5 to 2 Gy. Use of tangential fields
only resulted in more dose delivered to the surface of the opposite
breast, whereas use of the internal mammary field in addition to the
tangential portals gave more dose deeper in the breast; the range from
the lateral to the medial portions of the opposite breast was 1% to
4%. The volume of breast irradiated had minimal effect, but the use of
portals for the regional lymph nodes increased the dose to the
contralateral breast. Use of a 2.5-cm thick lead shield over the
contralateral breast during treatment with a medial tangential field
reduced the dose to 35% of its original value. Similar shields used on
the lateral tangent had essentially no protective effect. The authors
also recommended that wedges be used whenever possible on the lateral
tangential fields rather than on the medial to decrease the dose to
the contralateral breast.

A careful dosimetric study demonstrated that most of the scatter dose
received by the opposite breast originates in the collimator and
accessories of the accelerator, and it can be significantly decreased
by increasing the distance between the source and the patient's skin
(Fig. 50-39). [ref: 519] Therefore, an isocentric source-skin distance
technique may be desirable. Tercilla and associates, [ref: 762] using
thermoluminescent dosimetry in 15 patients treated with **60Co to the
breast and tangential fields, demonstrated that the contralateral
breast received 6% to 13% of the prescribed dose with the source-skin
distance (SSD) technique, compared with 4% to 9% with the isocentric
source-axis distance (SAD) technique. A greater contribution was given
from the medial tangential than from the lateral tangential beam. The
doses described with the SAD technique by these authors were higher
that those measured by Fraass and associates, [ref: 267] and they
believed the discrepancy could be attributed to the difference in the
treatment wedges used.

The use of half-field blocks (beam splitter) or, even better,
independent jaws combined with tailored beam splitters following the
contour of the chest wall of the patient, is very helpful in
decreasing the dose to the contralateral breast.

Yaparpalvi and colleagues, [ref: 863] in a study of nine patients
treated with breast conservation therapy and nine treated with
modified radical mastectomy followed by irradiation (with 6-MV photons
and tangential fields), used diode measurements and noted that the
scattered dose to the contralateral breast varied between 4.9% and
10.5% of the daily tumor dose. In addition, in patients receiving
supraclavicular or axillary irradiation with monoisocentric or
split-beam techniques, the scatter contribution to the contralateral
breast was about 1% of the prescribed dose; for a 50-Gy prescription,
this represented 2.47 to 5.3 Gy from the tangential fields and 0.5 Gy
from the supraclavicular and axillary fields.

Kelly and co-workers [ref: 388] reviewed the dose to the contralateral
breast from breast irradiation with tangential fields using four
different techniques. The highest dose was delivered with the use of
Cerrobend half-beam blocks (regardless of the proportion of wedge
used). Remaining techniques gave similar dose ranges, with the lowest
total dose produced by the asymmetric jaw with no medial edge
(Fig. 50-40).

It follows that attention should be focused on the medial beam in
attempting to reduce the contralateral breast dose. This could be
accomplished by treating without a wedge on the medial beam, but that
would compromise the dose distribution. Increasing the thickness of
the beam splitter on the medial field could lower the contralateral
breast dose, but the greater weight makes handling of the block by the
radiation therapist more difficult. The clinical significance of this
inadvertent radiation dose to the opposite breast is uncertain;
various investigators have failed to show an increased risk of
contralateral breast malignancy after treatment of the original breast
by radiation therapy. [ref: 20,80,508,641,801]

Irradiation of Regional Lymphatics

Supraclavicular Lymph Nodes

If only the apex of the axilla is treated (after modified radical
mastectomy or axillary dissection), the inferior border of the
supraclavicular field is the first or second intercostal space. The
medial border is 1 cm across the midline, extending upward, following
the medial border of the sternocleidomastoid muscle to the
thyrocricoid groove. The lateral border is a vertical line at the
level of the anterior axillary fold. The humeral head is blocked as
much as possible without compromising coverage of the high axillary
lymph nodes. This field is angled about 15 to 20 degrees laterally to
spare the spinal cord (Fig. 50-41A and Fig. 50-41B).

The low axilla is treated only when there is extracapsular tumor or if
axillary dissection is not performed. Here, the supraclavicular field
is modified: the inferior border comes down to split the second rib
(angle of Louis), and the lateral border is drawn to just block
fall-off across the skin of the anterior axillary fold.

The total dose delivered to the supraclavicular field is 46 Gy at 2 Gy
per day (calculated at a depth of 3 cm) in five fractions per week. An
alternative time-dose schedule is 50.4 Gy at 1.8 Gy per day.

Axillary Lymph Nodes

If the axilla is to be treated, the supraclavicular field is extended
to the second rib, as indicated previously. The dose to the midplane
of the axilla from the supraclavicular field is calculated at a point
approximately 2 cm inferior to the midportion of the clavicle. At the
end of the treatments to the supraclavicular field, the dose to the
midplane of the axilla is supplemented by a posterior axillary field,
as shown in Figure 50-42A and Figure 50-42B. The medial border of this
field is drawn to allow 1.5 to 2 cm of lung to show on the portal
film; the inferior border is at the same level as the inferior border
of the supraclavicular field; the lateral border just blocks fall-off
across the posterior axillary fold; the superior border splits the
clavicle, and the superior-lateral border shields or splits the
humeral head. Additional dose to the axilla midplane is administered
to complete 46 to 50 Gy (2 Gy daily). When indicated, a boost of 10 to
15 Gy is delivered with reduced portals.

Internal Mammary Lymph Nodes

The benefit of irradiation of the internal mammary lymph nodes is an
unresolved issue, since clinical failures at this site are very rare
[ref: 322] and most patients at risk receive some form of adjuvant
therapy. We agree with Lichter and associates [ref: 449] not to make
an effort to treat internal mammary nodes in most patients; we use
irradiation only in patients with primary tumors in the medial
quadrants that are larger than 3 cm in diameter.

The medial border of the internal mammary field is the midline; the
lateral border is 5 to 6 cm lateral to the midline; the superior
border abuts the inferior border of the supraclavicular field; and the
inferior border is at the xiphoid. If only the internal mammary nodes
are to be treated, the superior border of the field is at the superior
surface of the head of the clavicle. The field is set, as described,
with an oblique incidence to match the medial tangential portal. The
dose to the internal mammary field (45 to 50 Gy at 1.8 to 2 Gy per
day) is calculated at a point 4 to 5 cm beneath the skin surface.
Careful individualized planning and use of electrons of appropriate
energy for all or a major portion of the internal mammary node
irradiation are necessary to minimize dose to the lung. To spare
underlying lung, mediastinum, and spinal cord, electrons in the range
of 12 to 16 MeV are preferred for a portion of the treatment. The
usual proportion is 14.4 Gy delivered with 4- to 6-MV photons and 30.6
Gy with electrons (1.8 Gy daily).

Table 50-26 summarizes the recommended doses of radiation to be used
in combination with wide local excision of the primary breast tumor.

Timing of Irradiation After Conservation Surgery

The optimal sequence for combining breast-conserving surgery,
radiation therapy, and chemotherapy in the treatment of T1, T2, and
selected T3 breast cancers is not known and gives rise to heated
discussions. Recht and colleagues [ref: 593] reviewed the results in
295 patients who received breast irradiation and three or more cycles
of CMF or a doxorubicin-containing chemotherapy. The actuarial 5-year
breast failure rate was 4% in 99 patients who received irradiation
before chemotherapy, 8% in 54 patients who received initially some
chemotherapy or irradiation without concurrent chemotherapy followed
by further chemotherapy, and 6% in 116 patients who received
concurrent chemotherapy and irradiation. In contrast, the failure rate
in 26 patients who received all chemotherapy before irradiation was
41%. The actuarial 5-year local failure rate was 5% for 252 patients
irradiated within 16 weeks after surgery, compared with 35% for 34
patients irradiated more than 16 weeks after surgery.

Later, Recht and associates [ref: 594] reported on 244 women with
stage I or II breast cancer (209 node-positive, 35 node-negative) who
were randomly assigned to treatment with either initial radiation
therapy or initial chemotherapy; there were 122 women in each group.
Adjuvant chemotherapy consisted of methotrexate with leucovorin
rescue, 5-FU, cyclophosphamide, prednisone, and doxorubicin. The
proportion of women receiving at least 85% of the planned chemotherapy
dose was 46% in the irradiation-first arm and 64% in the
chemotherapy-first arm for cyclophosphamide and doxorubicin
(statistically significant); for methotrexate, the corresponding
figures were 22% and 46%. The crude local failure rates at 5 years
were 5% and 14% in the two arms, respectively, but the incidences of
distant metastasis were 32% and 20%, respectively. Table 50-27
illustrates different patterns of failure in the two treatment groups
as a function of various prognostic factors.

This study suggests that timing of irradiation or chemotherapy may be
decided on the basis of prognostic characteristics of the tumor:
patients with low metastatic risk should be treated with irradiation
first, and those with high metastatic risk should receive chemotherapy
before breast irradiation. Concurrent administration of both
modalities may be more appropriate, and ways to combine them without
sacrificing their effectiveness while minimizing treatment toxicity
should be explored in clinical trials.

Slotman and colleagues [ref: 703] reported on 508 patients with 514
stage I or II breast cancers treated with breast conservation therapy.
Patients with positive lymph nodes received adjuvant chemotherapy, and
postmenopausal patients received hormonal therapy. With a median
follow-up of 68 months, patients who started radiation therapy within
50 days of surgery had a recurrence rate of less than 2%, compared
with 6% for those with a longer interval. Forty-two patients started
irradiation within 25 days of surgery, and none of them had a breast
relapse.

Buchholz and associates, [ref: 82] in an analysis of 41 patients
treated with breast conservation therapy, reported that 25 who started
irradiation within 6 months after surgery had 100% local tumor control
at 8 years, in contrast to 80% control in 16 patients with delays of 6
months or longer (P = 0.02). The 8-year survival rates were 70% and
52%, respectively (P = 0.32).

Hartsell and associates [ref: 336] reported on 474 women treated with
breast conservation therapy, 84 of whom received chemotherapy for
positive axillary lymph nodes; 42 began receiving radiation therapy
before 120 days and 42 later than 120 days after surgery. With median
follow-up of 62 months, the 5-year actuarial breast relapse rate was
2% for those receiving early breast irradiation and 14% for those
receiving delayed irradiation (P = 0.05). Survival and disease-free
survival rates were not significantly different between the two
groups.

Nixon and colleagues [ref: 531] noted 5-year breast failure rates of
13% for 282 patients with an interval of 0 to 4 weeks between surgery
and breast irradiation (60 Gy to tumor bed) and 8% for 306 patients
with an interval of 5 to 8 weeks. The impact of the interval between
surgery and irradiation may be greater for times beyond 7 or 8 weeks.
At the Institut Gustave-Roussy, breast failure occurred in 17 (5%) of
361 patients irradiated within 7 weeks of conservation surgery,
compared with 7 (14%) of 51 patients who began radiation treatments at
7 weeks or later (P = 0.01). [ref: 118] However, this factor was not
statistically significant on multivariate analysis.

McCormick and co-workers [ref: 481] retrospectively analyzed 332
patients treated with conservation surgery and breast irradiation, 53
of whom received chemotherapy before irradiation and 86 of whom were
treated with a "sandwich" schedule (surgery, chemotherapy,
irradiation, chemotherapy). Median follow-up was 77 months in the
irradiation-alone group and 53 months in the chemotherapy-first
groups. Local tumor control was similar in all groups (96%, 86%, and
95%, respectively) at 60 months.

Buzdar and colleagues [ref: 86] retrospectively reviewed the records
of 552 patients treated with total mastectomy (467 patients) or
segmental mastectomy (85 patients) and irradiation plus combination
chemotherapy. A total of 463 patients received radiation therapy
first, and 89 received chemotherapy first. Median follow-up was 133
months for the total mastectomy subgroup and 39 months for the
segmental mastectomy subgroup. The incidence of locoregional failure
within each subgroup was not affected by the order in which
chemotherapy and irradiation were administered.

Leonard and co-workers [ref: 436] reported on 262 women in whom 264
breast cancers were treated, 105 with breast conservation therapy
including chemotherapy (group 1) and 157 with conservation surgery and
irradiation only (group 2). Eighty-five percent of the women received
all chemotherapy before irradiation, and 58% received hormone therapy.
There was no significant difference in the local recurrence rate by
time of initiation of radiation therapy, up to 6 months after local
tumor excision. The local recurrence rates were 4% in group 1 and 5%
in group 2.

The International Breast Cancer Study Group randomly assigned
premenopausal and perimenopausal patients with positive lymph nodes
and stage T1-T3,N1 breast cancer to receive either three or six
courses of CMF (trial VI) and postmenopausal patients to receive
tamoxifen for 5 years alone or combined with three cycles of CMF
(trial VII), both with or without three courses of delayed CMF. [ref:
828] Patients could be treated with either mastectomy or
breast-conserving surgery and irradiation. Radiation therapy was
delayed until the initial block of CMF was completed (4 or 7 months
after surgery for premenopausal and perimenopausal patients, or 2 to 4
months for postmenopausal patients). In both trials, 718 patients
elected to receive breast conservation therapy. The 4-year crude local
failure rates were 8% and 9% for premenopausal/perimenopausal patients
who had radiation therapy 4 or 7 months after surgery, and 3% and 6%
for postmenopausal patients who received radiation therapy 2 or 4
months after surgery. Therefore, in this study, moderate delays in
initiation of radiation therapy (4 to 7 months) did not adversely
affect local tumor control or survival in patients with invasive
breast cancer.

Reports on the impact of timing of irradiation and chemotherapy on
breast conservation therapy outcome are summarized in Table 50-28.
[ref: 600]

The influence of the interval between external irradiation and
brachytherapy was analyzed by Dubray and co-workers [ref: 176] in 398
breast adenocarcinomas (33 T1, 309 T2, and 56 T3) treated with
irradiation (45 Gy, 25 fractions, 5 weeks) followed by interstitial
**192Ir implant (37 Gy) to the tumor. Multivariate analysis showed an
increasing probability of local failure with longer interval between
external irradiation and brachytherapy (P = 0.005); there was a lower
risk of failure with complete tumor regression after external
irradiation (P = 0.022) and with a higher brachytherapy dose rate (P =
0.053).

Valero and Hortobagyi, [ref: 788] in an editorial, challenged reported
observations on the basis of lack of consistent data and strongly
recommended that patients treated outside clinical trials, especially
those with positive lymph nodes, receive adjuvant chemotherapy before
breast irradiation.

At present, it is generally agreed that irradiation should optimally
be started within 6 weeks after breast surgery for patients not
receiving chemotherapy and within 16 weeks for those treated with
adjuvant chemotherapy. Also, it is possible to administer two cycles
of CMF or even CAF chemotherapy (cyclophosphamide, doxorubicin, and
5-FU), followed by radiation therapy, and, if desired, to combine it
with cyclophosphamide and 5-FU (withholding the methotrexate or
doxorubicin), continuing three-agent chemotherapy after completion of
the radiation therapy. Another approach is to administer up to four
cycles of high-dose doxorubicin and cyclophosphamide after surgery,
followed by breast irradiation.

Preoperative Irradiation and Breast Conservation Therapy

An alternative to neoadjuvant chemotherapy is irradiation, although
the cytotoxic agents may have an additional benefit in eliminating
micrometastases. Calitchi and colleagues, [ref: 93] in 138 patients
with breast cancer unsuitable for initial primary tumorectomy, used
external breast irradiation (45 Gy in 25 fractions) to reduce the
tumor volume so that limited surgery could be performed (57% of
patients had tumor > 4.5 cm). After completion of irradiation, 22
patients (16%) showed no evidence of either clinical or radiographic
residual tumor and received an additional boost of 25 Gy. In 52 cases
(38%), tumor regression allowed tumorectomy followed by a boost of 20
Gy. Of the 64 patients with little or no tumor regression, radical
surgery was performed in 14 and high-dose curietherapy (37 Gy) in the
remaining 50, who refused mastectomy. Breast conservation therapy was
achieved in 74 patients (54%). The actuarial 5-year local tumor
control rate was 90%, and the disease-free survival rate was 73%.
Local tumor control in patients treated with brachytherapy was 76%,
and the 5-year disease-free survival rate was 65%. Of 59 patients who
underwent an axillary dissection, none had lymph node involvement.

Breast Conservation Therapy in Older Women

Fowble, [ref: 255] Maher and colleagues, [ref: 464] and Wazer and
associates [ref: 833] have suggested that less aggressive treatment in
elderly women with stage I or II breast cancer may not compromise
treatment outcome. Maher and colleagues [ref: 464] treated 70 women
with a median age of 81 years (range, 64 to 91 years) using
hypofractionated irradiation consisting of 6.5 Gy at weekly intervals
for a total dose of 45.5 Gy over 6 weeks; five fractions involved the
breast, and two fractions included the tumor bed. Patients were
treated in the lateral decubitus position to spare more rib and lung
than with tangential techniques. Tumor resection was not performed. If
regional lymph nodes were clinically involved, they were treated (27.5
to 30 Gy in five or six fractions). Tamoxifen (20 mg daily) was
administered. Twelve T1, 26 T2, 12 T3, and 19 T4 tumors were treated.
Forty patients (57%) had clinically negative lymph nodes, 25 (36%) had
N1, 2 (3%) had N2, and in 2 (3%) the nodal status was not recorded.
This course of radiation therapy was well tolerated by 87% of
patients. Overall survival at 3 years was 88%; disease-free survival
was 72%, and locoregional tumor control was 86%. Of 11 locoregional
failures, 8 occurred in the treated breast; 2 occurred simultaneously
in the breast and regional nodes, and 1 was in the lymph nodes. Late
complications consisted of moderate postirradiation fibrosis at the
tumor boost site in 39% of patients. No rib fractures, pneumonitis, or
brachial plexopathy was noted. Fowble [ref: 255] also noted that
elderly women tolerate breast irradiation as well as younger women.

Rostom and associates [ref: 627] also evaluated hypofractionated
irradiation (6.5 Gy once weekly for a total of six fractions) in
elderly patients with T1 or T2 breast cancer. With follow-up ranging
from 3 to 7.8 years, the local tumor control rate was 59% in 17
patients. Local tumor control was 90% in 6 patients with tumors
smaller than 5 cm who underwent excision before irradiation. Breast
fibrosis was observed in 22% of the long-term survivors, symptomatic
pneumonitis in 4 patients, and brachial plexopathy in 1. Fowble [ref:
255] cautioned about the potential for significant late effects with
high dose per fraction, because many of these elderly women have a
life expectancy longer than 10 years.

Wazer and associates [ref: 833] described results in 73 women aged 65
years or older (median age, 74 years) with stage I or II breast cancer
and clinically negative nodes who were treated with tumor excision,
breast irradiation (50 Gy at 1.8 to 2 Gy per fraction), a boost of up
to 20 Gy at the tumor-bearing quadrant, and regional lymph node
irradiation (45 to 50 Gy). Tamoxifen (10 mg twice daily) was given to
patients with positive hormone receptors. There were two breast
recurrences, yielding an 8-year local tumor control rate of 92.5%.
Regional nodal tumor control was 82.5% at 9 years. The 8-year
disease-free survival rate was 84%.


Patients treated with partial mastectomy and tamoxifen alone
experience a higher breast relapse rate than patients receiving the
same regimen plus irradiation. Cooke and associates [ref: 128]
identified 44 women who received partial mastectomy, breast
irradiation, and tamoxifen and compared them with 53 women treated in
a similar fashion but without breast irradiation. At 39 months, the
breast tumor recurrence rate was 5% with breast irradiation and 21%
when irradiation was omitted. All breast relapses occurred when
patients were receiving tamoxifen. Of those not receiving irradiation,
no breast relapses were seen in 22 patients older than 70 years of age
at diagnosis, in contrast to 8 breast recurrences in 31 patients
younger than 70 years.

Veronesi and associates, [ref: 803] in 299 patients with tumors less
than 2.5 cm in diameter who were treated with quadrantectomy combined
with axillary dissection and breast radiation therapy, reported a 0.3%
incidence of local recurrence, compared with 8.8% in 280 patients
treated with quadrantectomy and axillary dissection without immediate
irradiation. However, in women older than 55 years of age who did not
receive irradiation, the local recurrence rate was 3.8%. The 4-year
overall survival rate was similar in the two groups.

Therefore, elderly women with favorable prognostic factors may be
candidates for treatment with tumor resection and tamoxifen without
irradiation with close follow-up.

Follow-Up of Patients Treated with Conservation Surgery and
Irradiation

It is extremely important to closely monitor patients treated with
conservation surgery and irradiation, because early detection of a
local recurrence may allow for another wide local excision or a total
mastectomy, without significantly compromising the overall survival of
the patient. [ref: 259,331,407]

Breast clinical and self-examinations every 3 to 4 months for the
first 3 years, every 6 months through the fifth year, and yearly
thereafter should be emphasized. [ref: 596] The optimal interval for
follow-up mammography has not been determined. [ref: 263] In patients
with DCIS or invasive lesions, a baseline mammogram should be obtained
within 6 months of completion of treatment, and bilateral mammograms
should be obtained every 6 months or yearly for the first 2 or 3 years
(as dictated by findings) and yearly thereafter. [ref: 157,851]

If there is strong evidence of suspicious microcalcifications, masses,
or architectural distortions of the breast after conservation surgery
and irradiation, a biopsy should be obtained to rule out a recurrence.

At times, these patients are difficult to evaluate. Posttreatment
hematomas, fat necrosis, seromas, cysts, and scar tissue pose frequent
dilemmas. Consultation with an experienced mammographer is essential.
In 19 women treated with breast-conserving therapy,
microcalcifications were associated with the initial tumor in 5
patients. Eleven (58%) of the 19 biopsy specimens obtained for
suspicious posttreatment findings were positive for recurrent breast
cancer (4 invasive ductal, 2 microinvasive ductal, 4 intraductal, and
1 lobular carcinoma in situ). [ref: 716] If there is doubt, repeat
biopsy may be indicated.

The way in which local recurrence after breast-conserving treatment
for invasive carcinoma became apparent was reported in 44 patients
monitored with regular physical examination and annual mammography.
[ref: 631] In 36 patients (81%), the first suspicion of local
recurrence was heralded by clinical signs and symptoms; they presented
between two scheduled routine visits in 12 patients, at the time of a
routine visit in 14 patients, and at routine physical examination by a
surgeon or radiation oncologist in 10 patients. Mammography detected
the recurrence in only 8 patients, and it confirmed the clinical
suspicion in 7 patients. The remaining 23 patients with clinical overt
recurrence showed no signs of recurrent tumor on the mammograms
performed after first clinical suspicion. Fine-needle aspiration
cytology confirmed the clinical suspicion in 35 of 38 patients. In the
authors' experience, regular physical examination is the mainstay for
the detection of local recurrence after breast-conserving therapy.
Mammography was of limited value but proved more valuable for the
early detection of recurrent tumor outside the primary tumor area.

Pezner and colleagues [ref: 569] evaluated results of 38 posttreatment
biopsies (with benign pathologic findings) that were performed on 32
irradiated breasts or axillae in 31 of 232 patients who underwent
conservation treatment of early-stage breast cancer. Postbiopsy
wound-healing complications developed in 8 (30%) of 27 patients who
underwent open biopsies but in none of 11 who had only needle
biopsies. Wound-healing complications occurred in 2 of 5 patients who
underwent incisional skin biopsy, 3 of 5 who underwent mammographic
needle-localized excisional biopsy, and 3 of 17 with other types of
open biopsies. Wound-healing delay occurred in 4 of 6 patients with
larger breasts, compared with only 4 (19%) of 21 patients with smaller
breasts. Significant worsening of cosmetic breast retraction was
frequently associated with wound-healing complications.

Recht and Harris [ref: 598] reported on 1233 patients with clinical
stage I or II carcinoma of the breast treated with excisional biopsy
and irradiation who were monitored periodically; 38 had pathologically
negative posttreatment ipsilateral breast biopsies at different times
(range, 5 to 59 months; median, 26 months). The most frequent findings
just before the first biopsy were palpable mass with or without
induration in 15 patients (39%) and thickening of the breast or
fullness without a mass in 20 patients (53%). Only 21 patients had
mammograms within 4 months before the biopsy; 3 showed a mass only
(14%), 4 had breast findings without a mass (19%), 1 had both
findings, and 11 (52%) had no suspicious findings. Two patients (5%)
undergoing biopsies later developed breast failures.

Therefore, both periodic careful physical examination and mammography
are critical in the posttreatment evaluation of patients treated with
breast conservation therapy. At least yearly evaluation is mandatory
even 10 years after therapy because of the possibility of late breast
relapses and occasional distant metastases.

Radiographic Findings After Breast Conservation Therapy

Dershaw [ref: 157] summarized the most frequent mammographic findings:
parenchymal distortion and fibrosis at the tumor excision site
(secondary to surgical scar and irradiation); skin thickening, seen in
90% of patients, may be diffuse or more prominent at the surgical
excision site; calcifications, due to fat necrosis, are coarse and
round and have radiolucent centers. If these findings are stable,
mammographic follow-up is sufficient; however, a change in number or
characteristic pattern warrants a biopsy to rule out recurrent tumor.
Mammographic findings were correlated with clinical observations in
several studies. [ref: 76,405,493] Figure 50-43 demonstrates
mammographically observed changes occurring in a patient who were
treated with breast conservation therapy. These studies demonstrated
that most changes were observed in the first 12 months after therapy,
with stabilization achieved at 12 to 36 months after completion of
therapy. Breast edema is mammographically present in virtually all
patients at completion of therapy, with a steady decline in this
observation over 36 months and approximate stabilization at 42 months
(Figure 50-44).

Preoperative and postoperative mammograms were reviewed in 103
patients undergoing conservation therapy. [ref: 76] The main
posttreatment findings were diffuse increase in parenchymal density
with coarse stromal pattern, some parenchymal distortion, and
thickening of the skin. Changes reached a peak at 9 months and slowly
resolved over the next 2 years. At 31 to 33 months, 3 of 15 patients
still had dense parenchyma, and 6 had skin reaction. Sixty-nine
patients had heavy fibroadenosis. Scar with retraction in the surgical
area was observed on the mammograms of 71 patients. Fat necrosis was
noted in 2 patients. During the 3-year follow-up, recurrent cancer was
noted in 2 treated breasts, and 3 women developed contralateral breast
cancer.

Orel and colleagues [ref: 539] reported on 1145 women with early
breast cancer treated with lumpectomy and irradiation; 102 women with
various mammographic and clinical findings later required biopsy at
the treated site, and 58 had two sets of mammograms available for
review (1 within 3 months of the biopsy). Recurring cancer was
documented in 38 (66%) of 58 patients; 13 (34%) of the recurrences
were detected solely with mammography, and 8 others were detected both
mammographically and clinically. The positive predictive value for
mammographic abnormalities was 72% (76% for soft tissue
microcalcifications and 62% for other findings). Twenty-one
recurrences (55%) were within the lumpectomy quadrant. Within the
lumpectomy site, sensitivity was substantially better for physical
examination (71%) than for mammography (43%). In the remaining breast
outside the lumpectomy quadrant, mammography had a significantly
higher sensitivity (71%) and positive predictive value (86%). The most
common posttreatment findings reported by Orel and colleagues and by
Stomper and associates [ref: 735] were calcifications alone or with a
mass distortion of the breast parenchyma, and inflammatory thickening
of the breast skin (Table 50-29). In an update of their experience,
Orel and co-workers [ref: 538] reported on 72 women who had
ipsilateral breast recurrences after breast conservation therapy and
were treated with salvage mastectomy. In 24 patients (71%)
calcifications were observed on the mammogram, and 2 patients (14%)
had both calcifications and an abnormal physical examination. In 9
patients (26%), a mass was thought to be significant on mammography,
and 12 (86%) had positive physical examinations. The postrecurrence
5-year survival rates were 88% for patients with recurrent tumor
detected by physical examination with or without abnormal mammograms
and 73% when recurrences were detected by mammography only (P > 0.21).

Stomper and colleagues [ref: 735] reported on 50 of 1600 patients with
stage I or II invasive breast cancer treated with conservation surgery
and irradiation on whom biopsies were performed within 4 months of a
mammogram for suspected recurrence in the irradiated breast. The tumor
was suspected on mammogram in only 8 patients (35%), on physical
examination in 9 (39%), and on both in 6 (26%). The most common
radiographic findings were calcifications with or without a mass.
Histologic evidence of recurrent cancer was found in 23 (51%) of 45
biopsy specimens. Sixty-five percent of patients had recurrences at
the primary site and 22% in other sites; 13% were multifocal.

Dao and associates [ref: 144] evaluated 35 women with breast carcinoma
treated with conservation therapy who underwent posttreatment magnetic
resonance imaging. Nine patients had recurrent tumors, and 26 had a
benign fibrotic mass confirmed at biopsy. In all cases, a localized
hypointense area was present on plain spin-echo T1-weighted images. In
all recurrent tumors, dynamic gadolinium-enhanced T1-weighted images
demonstrated early increased signal intensity of the lesion within 3
minutes after bolus injection. The signal intensity over time in
localized fibrosis differed from that in tumor recurrence, with no
substantial enhancement on postcontrast T1-weighted images.

It is important to define the cost-benefit of follow-up procedures. In
a controlled trial in Italy, 655 women were randomly assigned to be
monitored with an intensive surveillance program including physician
visits, bone scan, liver ultrasound, chest x-ray, and laboratory tests
after initial treatment for breast cancer. [ref: 291] A control group
of 665 women were monitored by their physicians with physical
examination and only clinically indicated tests. Both groups received
a yearly mammogram. Compliance in both protocols was more than 80%.
With a median follow-up of 71 months, there was no difference in
overall survival between the two groups. There were 132 deaths (20%)
in the intensive surveillance group and 122 deaths (18%) in the
control group. Time to detection of recurrence and parameters related
to quality of life were equivalent in both groups. Therefore,
unnecessary tests are discouraged in the follow-up of patients treated
for breast cancer.

Quality Assurance in Breast Conservation Therapy

In terms of quality control, the surgeon carries responsibility for
the accurate measurement and recording of clinical signs at the time
of first presentation. Furthermore, the surgeon must obtain an
adequate histopathologic specimen of the primary tumor. If technically
feasible, a complete tumor excision should be accomplished with
sufficient margins. Reexcision is of value in achieving negative
margins. The specimen should be appropriately marked (with sutures)
for orientation. The pathologist must process the specimen in
accordance with established guidelines, inking the margins and
obtaining all necessary sections to assess the status of all margins.
Sufficient tissue should be secured, whenever possible, for
quantitative determination of hormonal receptors. Guidelines for
evaluation and reporting of pathologic features have been published.
[ref: 653] The surgeon must formulate a plan for the most effective
treatment of the primary tumor and axillary lymph nodes. As part of
quality control in management of invasive carcinomas, axillary
clearance should be intrinsic to the primary treatment. Failure to
clear the axilla can result in understaging; as a result, adjuvant
systemic therapy may not be given to patients who could benefit from
such treatment. There is need for close cooperation among surgeons,
diagnostic radiologists, pathologists, and radiation and medical
oncologists in the management and follow-up of patients with early
breast cancer. [ref: 208]

A consensus on a quality assurance program for the treatment of early
breast cancer in Europe included guidelines for treatment preparation
and execution, careful location and excision with marking of the
primary tumor, target volume definition for irradiation of the whole
breast and boost area, irradiation dose prescription, specification
and checking procedures, and measures to achieve a homogeneous dose
within the target volume. [ref: 42]

Accuracy in the dose specification of breast irradiation techniques
was evaluated at participating institutions in an EORTC trial. [ref:
796] Three transverse sections of a patient were sent to 16
participating institutions with a request to make a three-plane
treatment plan according to the protocol prescriptions. A treatment
chart and beam data were also requested for recalculation of the dose.
Recalculation of the dose at the isocenter showed agreement within 2%
of the stated dose. The dose reported in the tumor excision area
varied between 93% and 100%.