BREAST IMAGING : CURRE'.JT STATUS AND FUTURE DIRECTIONS 0033-8389/92 $0.00 + .

20

EVALUATION OF THE
POSTOPERATIVE BREAST
Ellen B. Mendelson, MD

The large number of surgical procedures LUMPECTOMY/BENIGN

involving the breast can be extrapolated from BIOPSY
the expected number of cancers.1 2 About
175,000 new breast carcinomas will be diag­ Background, Eligibility, and
nosed during the current year in the United Role of Mammography in
States. If this number represents approxi­ Patient Selection
mately 15% to 35% of surgical breast biopsies,
In the last tw o decades, breast cancer ther­
the total number of surgical procedures
apy has changed dramatically. For eligible
would be at least 500,000. Because postsur­
women, equivalent survival rates have been
gical changes may overlap with radiographic
demonstrated for breast conservation ther­
features of malignancy, mammographic eval­
apy (wide tumor excision and radiation ther­
uation in these patients may be difficult. apy) and mastectomy 29 Tumor recurrence
Alterations in breast tissue also may change varies from 6% to 10 %, at rates reported as
in time, further complicating interpretation. 1 % to 2% or more per year after treatment. 5.25.
For analysis to be accurate, the mammo­ 29. 36. 83. 86 In the first 7 years, the tumor
grams must be placed in temporal context generally recurs near the lumpectomy site,
and correlated with the physical findings and with mean time to recurrence being 3 years.
procedures that have been performed. A fa­ After that, tumor is found increasingly in
vorable outcome for a breast cancer patient other quadrants. The conservatively treated
requires the combined efforts of the radiolo­ breast cancer patient with recurrent local tu­
gist, pathologist, surgeon, radiation oncolo­ mor does not have the same poor prognosis
gist, medical oncologist, and the patient her­ as a patient with a mastectomy who has a
self. Particularly when the carcinoma is chest wall recurrence. Salvage mastectomy
nonpalpable, the radiologist plays a central to treat recurrence after lumpectomy and
role in the team effort. radiation therapy does not jeopardize a pa­
This article focuses on imaging and follow­ tient's survival expectations. In some cases,
up management of the conservatively treated the recurrent carcinoma has been reexcised
breast. Also reviewed are the radiographic without sacrificing the breast.
findings associated with benign breast biopsy Careful selection and staging of patients
and cosmetic procedures including recon­ for breast conservation therapy are important
struction, augmentation, and reduction for optimal outcomes. 29 . 68. 88 Although eligi­
mammoplasty. bility criteria may vary somewhat, candidates

From the Department of Radi ology, The Wes tern Pennsy lvani a Hospita l and the Uni versity of Pittsburgh School of
Medicine, Pittsburgh, Penn sylva nia

RADIOLOGIC CLIj\;ICS OF NORTH AMERICA

VOLLv1E 30 • NUMBER 1 • JANUARY 1992 107

108 MENDELSON

for lumpectomy and irradiation will have microcalcifications better and exclude benign
tumors less than 5 cm (Tl or T2), although calcifications such as milk of calcium in small
most tumors have been smaller than 4 to 4.5 cysts (90-degree lateral view). Clusters of
cm. 29 . 62 Positive axillary lymph nodes are not dermal calcifications can be identified with
a contraindication . An important selection magnification views that might show their
criterion is that the tumor be removed with radiolucent centers and tangential views for
satisfactory cosmesis. Excision of a large tu­ localization in the skin. Magnification radiog­
mor from a small breast may result in breast raphy is also useful to bring out other faintly
deformity and a poor cosmetic effect. ;'\;0 seen foci of involvement and to evaluate the
location in the breast is absolutely contrain­ retroareolar area. Many more microcalcifica­
dicated for breast conservation therapy. Ret­ tions may be present than seen on standard
roareolar lesions involve removal of the nip­ projections. 73
ple areolar complex, and for some patients, A factor that may affect risk of recurrence
mastectomy may be preferable. Women with is size of the tumor, a good prognosis for
multicentric masses or diffuse, widespread tumors smaller than 1 cm. 32, ,·2 If the tumor
malignant appearing microcalcifications are can be outlined on a mammogram or seen
not good candidates for lumpectomy and on a sonogram, three dimensions can be
radiation therapy 62. 79. 80 measured. Spot compression, with or with­
Breast conservation therapy that requires out magnification, can be used to spread
irradiation is not an option for patients who apart overlapping areas of breast tissue, ex­
are in the first and second trimesters of clude a pseudomass caused by superimposed
pregnancy. 38 If lumpectomy is performed in parenchyma, and image marginal detail of a
the third trimester, the breast carcinoma can true mass. Spot compression should be used
be irradiated after deliverv. Women who to define the margins of discrete, rounded
have collagen vascular dise~se are at risk of soft-tissue densities that might represent tu­
breast fibrosis after radiation therapy. 36. 38. 7! mor satellites near the tumor or multicentric
In patients who have had radiation therapy foci in distant quadrants.
previously to an area that has included the Presurgical localization should be precise,
breast, such as for Hodgkin's disease, irra­ usually no more than 0.5 to 1.0 cm away
diation for breast carcinoma results in an from the mass or calcifications. 33 . n 86
unacceptably high cumulative dose . Breast Whether the level of suspicion of malignancy
conservation therapy is not a good choice for is high or low, all presurgically localized
debilitated patients for whom the long com­ non palpable and some palpable abnormali­
mitment to therapy would be a hardship. ties require specimen radiography to confirm
Those women who do not wish to undergo removal. If dense fibroglandular tissue ob­
lumpectomy and radiation therapy should be scures the mass or microcalcifications in the
offered other treatments. specimen, a different projection may shift
tissue relationships and permit the abnor­
mality to be perceived.! ! Full compression
Preoperative and and magnification also increase visibility of
Perioperative Assessment: the abnormality. If the mass is present at one
The Role of Mammography or more margins or microcalcifications extend
to the edge of the specimen, the surgeon,
When tumor is demonstrable radiographi­ before closure, should be advised to remove
cally, mammography can establish appropri­ more tissue. Additional excised material
ateness of lumpectomy and radiation therapy should also be radiographed. Each specimen
by defining extent of a patient's disease, should be numbered, correlated with the
suggesting multicentricity, and evaluating mammogram, and described, with findings
the contralateral breast?4 Mammographic conveyed by the radiologist immediately to
analysis can further help to guide patient the surgeon in the operating room and then
selection by suggesting presence of certain later in the written report.
prognostic indicators for tumor recurrence Perioperative radiographic assessment of
such as invasive carcinoma with an extensive the adequacy of resection is one of the ra­
intraductal component as manifested by mi­ diologist's major responsibilities. If removal
crocalcifications seen on the mammogram. 40 · 6 1 of the abnormality remains uncertain after
Magnification radiography should be per­ specimen radiography and additional tissue
formed in two projections to characterize sampling, intraoperative mammography has
 EVALuATION OF THE POSTOPERATIVE BREAST 109

been suggested if local anesthesia has been the breast tissue to reach the lesion from a
given. 84 After loose suturing of the incision, circumareolar incision is acceptable for a bet­
a mammogram is performed. If the lesion is ter cosmetic result. 16 Tunneling is avoided
confirmed in the breast, the abnormality is otherwise in the breast cancer patient be­
then relocalized and the patient returned to cause of difficulties in radiation treatment
surgery. Merits of the procedure have occa­ planning and follow-up care where the skin
sioned some debate, calling attention to the incision site and the deeper tumor bed do
radiologist's central role in patient manage­ not correlate.
ment. 35 Cosmesis is one of the requirements of
successful breast conservation therapy, and
the surgical technique should aspire to a
BREAST CONSERVATION IN good cosmetic result as well as control of
CARCINOMA disease. 2u8 Ordinarily, in a lumpectomy only
the subcutaneous fat and subcuticular layers
are sutured. The deeper tissues of the sur­
Surgical Approach gical bed fill in gradually. Meticulous hemo­
stasis decreases the likelihood that large he­
In the United States, the surgical approach matomas or seromas will form in the biopsy
for excisional biopsies is the same as that for site. Cse of a drain is discouraged. lb , 29 The­
lumpectomy for carcinoma. 82 Through a cur­ oretically, using this surgical technique, the
vilinear incision made directly over it, the normal breast contour might be preserved
lesion is removed, surrounded by a rim of better than where apposition of breast paren­
grossly normal breast tissue (Fig. 1).29,30,79,92 chyma and the placement of a drain could
Ordinarily, no skin is taken with the lesion; create a craterlike concavity.
if any skin is excised, it should be only a Terms for the surgical procedures of breast
small ellipse. Unless the tumor is adherent conservation therapy are not defined consis­
to the pectoralis fascia, the fascia deep to the tently." For Kinne and Kopans, lumpectomy
tumor is spared. In a young patient in whom and excisional biopsy signify tumor removal
an abnormality is thought to be benign, such without regard to marginal status.;~ They
as a fibroadenoma, some tunneling through distinguish these synonyms from wide exci­
sional biopsy or removal of the tumor with a
surrounding area that is free of malignancy
grossly or histologically. In the literature, the
terms excisional biopsy, wide excision, tumorec­
to my, lumpectomy, segmental mastectomy, and
tylectomy are often interchanged. A descrip­
tion of the procedure is ordinarily provided
in each reference.
The wide excisional biopsy (partial mastec­
tomy, limited resection) is the definitive sur­
gical procedure for conservative treatment of
\ breast carcinoma in the Cnited States. 4" Other
surgical techniques, such as a quadrantec­
tomy, are advocated in Europe for tumor
removal, with removal of the quadrant of
breast tissue containing the tumor along with
the skin and superficial pectoralis fascia 90, 91
Except where confirmation of carcinoma has
been provided by fine-needle aspiration or
core biopsy in advance of surgery, additional
surgical procedures await histologic interpre­
Figure 1. Breast conservation therapy: surgical technique.
tation. Thus, axillary dissection or mastec­
Through an incision made directly over the tumor, the tomy may be performed as the second of a
tumor is removed with a surrounding area of grossly two-stage procedure. 16 Other theories of tu­
normal parenchyma. A separate incision is made for the mor spread may lead to different approaches
axillary dissection (arrow), which may be performed later.
(Adapted from Isaacs JH: Breast biopsy and the surgical
to resection such as excision of a discharging
treatment of early carcinoma of the breast. Obstet Gynecol ductal segment mapped preoperatively with
Clin North Am 14:711-732, 1987.) galactography.
 110 MENDELSON

Radiation Therapy Accurate interpretation depend s as much

on the chronology of altera tions as it does on
The affected breast is ordinarily irradiated mammographic depiction of masses and cal­
as soon as the surgical site has h ealed ade­ cifications on each single examination. 15. 40
qua tely, 2 to 5 weeks after lumpectomy an d \!Iany errors will be avoided if mammograms
axillary dissection. Forty-five to 50 Gy is are evaluated in sequence, always comparing
given to the breast in five daily doses per with th e earliest not jus t the most recent
week over a 5-week period. 3~ se For many study. If eva luated in this manner, the pos­
patients, an electron beam or iridium implant sibility of recurrent carcinoma can be sug­
boost to the lumpectomy site will be used, gested with greater confidence. Misinterpre­
increasing the total dose to the breast to tation of surg ical and radiation changes also
approximately 60 to 66 Gy. After lymph node will be less likely, which is important in
dissection , the axilla is not irradia ted. 62 avoiding unn ecessary biopsy of rad iated tis­
sues, which may heal less readily than in the
untreated breast.
IMAGING AFTER BENIGN
BIOPSY AND LUMPECTOMY
Patient Data
AND RADIATION THERAPY
For the approp ri a te examination to be per­
The purposes of breast imaging after sur­ formed , the radiologist and technologist
gical biopsy are to confirm removal of the must be aware of pertinent aspects of the
abnormality, to assess postprocedural com­ patient's history. A diagram and checklist of
plications, to detect recurrent tumor at the signs and symptoms can be printed on the
operative site, and to demonstrate other ip­ referring physician' s prescription pad used
silatera l or con tra la teral interval changes that to request a mammogram. In many practices,
mig ht signi fy carcinoma. Expected postop­ including our own, the patient is give n a data
era tive changes include ma sses, scarring, sheet to complete. In addition to ques tions
edema, skin thi ckening, and calcifi cations. * related to family history and other possible
These postoperative find ings, which may ri sk factors for breast carcin oma, the patient
mask the signs of malignancy, occur both is asked if and when she ha s had surgical or
after benign biopsy and excision of carcinoma other treatment for breast cancer, biopsies,
and are accentuated and prolonged by radia ­ aspirations of flui d-filled or solid lesions, or
tion therapy. cosmetic alterations. The patient is requested
to indicate locations of masses , pain, dnd
prior surgery on a diagram. On subsequent
General Approach visits, the patient will fill out an abbreviated
form to update her history of surgical pro­
Diagnostic accuracy increases with the cedures a nd medications, including post­
awareness (1) of how the procedures are menopausal hormonal replacement therapy.
performed, (2) that there ma y be drastic or Th ese informational form s and copies of all
little temporal change, and (3) that interpre­ pathology and cytology reports are kept in
tation is even more difficult in the densely the patient's folder \-vhere they are easily
fibroglandular breast 3 -1 accessible for patient management decisions
The potential of mammograph y to charac­ and case review conferences.
terize physical and mammographic findings
should be fully utilized. Additiona l views,
such as spot compression, ma gnification, Marking Scars
tangential , and various obliquities w ill be
useful in mos t cases. To identify the fluid Technologists take a brief history from the
component of a postoperative mass, sonog­ patient. They mark the location of palpable
raphy is indicated,I-6 and other techniques masses, dermal lesions, and scars on a dia­
such as compu ted tomography (CT) or mag­ gram and then on th e patient. Small radi­
netic resonance (MR) imaging will be helpful opaque BBs are placed on the site of dermal
occasionally.42. -13. S6 lesions or palpable masses. For correlation
with possible mammographic findings such
"References 6, 7, 10, 22, 23, 34, 39, 40, 51, 52, 55, 56, as architectural distortion and new dys­
59, 65-67, 73, 74, 76, 86, 89. trophic calcifications, scars are marked with
 EVALCATION OF THE POSTOPERATIVE BREAST 111

lengths of thin wire that are taped to the hypoechoic area with posterior acoustic shad­
skin. Wires that are used for presurgical owing (Fig. 3) seen when scarring has devel­
localization of nonpalpabJe lesions are of the oped. 56 :'--Jot infrequently, a tract between the
appropriate caliber, and they can be made surgical bed and the skin incision site can be
easily to conform to the length and shape of demonstrated.
the scar. Fine-gauge radiopaque metallic Some surgeons deploy clips within the
wires are also obtainable inexpensively from breast at the margins of the lumpectomy site
jewelry and craft shops. A piece of wire the to focus mammographic follow-up and for
size of the scar is affixed to the adherent radiation therapy planning (Fig. 4). The in­
surface of hypoallergenic paper tape, which terclip distance provides a measurable refer­
is then placed on the scar before the mam­ ence for the surrounding soft-tissue density,
mogram is performed. which represents postsurgical fibrosis and fat
Although the breast may show no radio­ necrosis. On follow-up studies, there will be
graphic evidence of prior surgical activity, contraction of the scar tissue. Stability is
particularly with benign biopsies after several defined as lack of interval change on two
years have passed (Fig. 2)/6 application of a successive studies. After that, a new nodule,
skin marker to the patient adds little to the microcalcifications, or an increased area of
time required for the examination. Marking soft-tissue density surrounding the clips or
all scars will ensure that the maneuver will even separating them will suggest recurrent
not be forgotten in the conservatively treated tumor.
breast cancer patient in whom mammo­ Neither the surgical clips nor the skin wires
graphic changes may be more profound, per­ will interfere with imaging. Except to cause
sistent, and ambiguous. An exception is a small signal void artifact, the clips should
made for extensive surgical procedures, such not interfere with :-vIR imaging, a technique
as reduction mammoplasty, that involve being studied for its potential to allow differ­
much of the breast, which are shown dia­ entiation of mature postsurgical scarring
grammatically only. In demarcating the scar, from recurrent tumor at the lumpectomy
the marker reminds the technologist to po­ site. 43 . 44
sition the patient carefully so that the entire
area is included on at least one view. The
marker also can be used to direct patient
positioning for tangential views that will de­ FINDINGS AFTER BENIGN
fine the relationships of a mass or calcifica­ BIOPSY AND LUMPECTOMY
tions to the skin. Mammographic visualiza­ AND RADIATION THERAPY
tion of the breast tissue is not compromised
by the wires. Masses and Fluid Collections

Asymmetric soft-tissue densities are ex­

Identifying the Surgical Bed pected at postoperative sites. 85 They may
represent fluid collections, fibrosis, or fat
It is important to evaluate the tumor bed necrosis at the site of surgical activity. Paren­
because more than 65% of recurrences are in chymal asymmetry may also come to atten­
or within a few centimeters of the cite of tion as a possible contralateral abnormality,
excision. 32 . 37. 38. 87 The skin marker on the scar ultimately explained by absence of equivalent
may not always correlate with the site of tissue on the operative side (Fig. 5). The
surgical excision lying deeper in the breast. appearance of the surgical site depends upon
A mammographic view tangential to the skin the interval that has elapsed between the
incision site will permit differentiation of an procedure and the imaging studies. During
external scar from the spiculated intramam­ the first year after breast conservation ther­
mary scar. apy, in the area of lumpectomy, the mam­
Ultrasonography can be used similarly to mogram often shows an oval mass that is
distinguish the skin from the lumpectomy fairly dense and well marginated but usually
scar within the breast. On posttreatment with a few spiculations or irregularities. On
sonograms, the surgical bed can be identified a 90-degree lateral view, fluid elements
beneath the subcutaneous fat as a distinct within these hematomas or seromas may
hypoechoic oval area with posterior acoustic show layering. 23 The mammographic appear­
enhancement if fluid remains or as a linear ance and timing are suggestive; tumor re­

r-"
Figure 2. Resolution of postsurgical change after benign biopsy. A. In a patient who already had been biopsied in the
inner half of the breast (note wire on skin), the two soft-tissue densities were localized and excised. The smaller, medial
mass was a fibroadenoma and the larger, a focal area of chronic cystic disease. No specimens were submitted for
radiography at the time of biopsy. B, The diagnosis of the larger mass was questioned, and a postoperative mammogram
was requested. Two months following biopsy, there are large spiculated masses and mild edema. Detection of residual
mass is not possible with this study, and follow-up was suggested. C, Six months later, 8 months after the biopsy, the
biopsy site in the outer half of the breast has healed nearly to completion, and there is considerable resolution at the
medial site. Postsurgical architectural distortion remains and edema has cleared. D, Two years after the biopsy, little
evidence of surgical activity is seen.
112
 EVALUATION OF THE POSTOPERATIVE BREAST 113

Figure 4. Surgical clips demarcating lumpectomy site.

Surgical clips outlining the site of tumor removal are useful
markers for radiation therapy planning and mammo­
graphic followup studies. (Courtesy of David M. Van Hook,
MD.)

growth to larger than preoperative dimen­

sions is unlikely within the first year. If there
is doubt, sonography is indicated to identify
the fluid-filled nature of the mass. Some
hematomas have echoes within them ini­
tiallv,l but most soon become anechoic. Post­
lum'pectomy collections show posterior
acoustic enhancement. The location, shape,
and some marginal irregularity correlate with
the surgical excision. Septa may be present
and do not signify complications. ' 6
Management of a complex mass requires
knowledge of the clinical context. Clumps of
echogenic material may represent thrombus
Figure 3. Postoperative scarring. A, Postoperative pre­ and abscess formation is uncommon. If the
radiation mammogram shows asymmetric area of soft­ fluid-filled region is not tense or painful and
tissue density and architectural distortion (arrows) at the
site of excision of an infiltrating ductal carcinoma 4 weeks
if abscess is not suspected, observation may
earlier. B, Correlative sonogram demonstrates the shape, be preferred to either aspiration or incision
depth, and size of the parenchymal scar (arrows) in the and drainage of tissue in which healing
breast, with a tract (thin arrow) connecting it to the mechanisms may be somewhat compromised
thickened skin (short arrows) at the incision. A small by radiation therapy.
amount of residual fluid manifested by posterior acoustic
enhancement is present in the surgical bed with fibrosis
Postsurgical fluid collections are also seen
developing in the deeper, irregularly marginated, shad­ following benign biopsies. If the specimen
owing portion . radiograph has confirmed removal of the
114 :vIENDELSON

initially were still present at 6 months, with

20 % demonstrated at 9 months. Ninety-six
percent of fluid collections had lost the ultra­
sonographic features of cysts by 1 year, and
nearly 100 % by 18 months.
The percentage of patients with postsur­
gical fluid collections will undoubtedly reflect
the surgical technique. In the group of pa­
tients we studied, tumor excision was per­
formed as advocated by Fisher,29 that is with
only subcuticular and skin closures and with­
ou t drainage of the site. Theoretically, grad­
ual fluid resorption will result in a more
normal breast contour, but a few surgeons
prefer to aspirate or drain these fluid accu­
mulations briefly. Advantages of evacuation
of the fluid would be to accelerate scar for­
mation and minimize the already low inci­
dence of abscess formation at the surgical
site. For the radiologist, a hematoma or ser­
oma obscures the surgical bed, and aspiration
allows better breast compression and pene­
tration of the lumpectomy site for the mam­
mogram .
As the lumpectomy site is imaged on fol­
low-up mammograms during the next 6 to
18 months, the discrete, dense, fairly well­
Figure 5. Asymmetric density after benign biopsy. rv1edio­ marginated mass that represents the fluid
lateral oblique projections of the breasts show an area of
soft-tissue density in the right upper breast. After exclusion
collection will begin to diminish. As scar
of a mass with other radiographic projections, the asym­ formation progresses, areas of radiolucency
metry is ascribed to removal of similar tissue from the left are seen interspersed with the soft-tissue
upper breast during a previous biopsy. For correlation density. The radiolucencies represent fat en­
with changes, the biopsy incision is marked by a thin wire
taped to the skin. No scarring or architectural distortion
trapped by the developing scar. As fluid is
remains in the left upper breast after this patient's biopsy. resorbed, the lesion may elongate and be­
come poorly marginated and spiculated .
Sonography performed between 6 and 12
abnormality and the pathology report does months may show a complex mass contain­
not suggest the need for urgent imaging, the ing a well-margina ted resid ual cystic portion
patient will be placed in a routine follow-up and a component with posterior acoustic
category. At 1 year, architectural distortion shadowing suggesting formation of fibrosis
or a spiculated soft-tissue density represent­ (Fig. 6). When scar evolution is complete, by
ing an evolving scar may be seen. Few post­ 12 to 18 months, the cystic areas will no
surgical collections remain 1 year after benign longer be seen.
biopsy. 56. 76
Approximately half of 110 breast cancer
patients we studied 4 weeks after lumpec­ Asymmetric Soft-Tissue
tomy and prior to radiation therapy had fluid Density with Architectural
collections at the surgical site. Most often, Distortion or Spiculation
their size was 3 to 5 cm in longest dimension,
but occasionally collections were larger. In Thickened skin at the incision should be
our experience, the sonographic characteris­ distinguished from the parenchymal scar of
tics of cystic lesions are retained much longer tumor removal. Prominent scarring in the
than the 2 to 3 weeks reported by Peters66 breast develops in more than 95S{ of patients
and 2 to 8 weeks by Sadowsky.74 Paulus") by the end of the first year after lumpectomy
noted persistence of these fluid accumula­ and radiation therapy. Benign biopsy
tions for months and, in some cases, years. changes often resolve more quickly and com­
Fifty percent of fluid collections we identified pletely 76 Because both scarring and carci­
 EVALCATIO N OF THE POSTOPERATIVE BREAST 115

Figure 6. Resolving fluid collection after lumpectomy and radiation therapy. A, Small oval fluid collection 6 months after
breast conservation therapy. B, At 1 year, the skin is still thick. A small residual fluid collection (F) is present centrally,
and scarring is developing at the periphery where posterior acoustic shadowing (5) is seen. Ultimately, an irregular
hypoechoic shadowing area will be present.

noma are spiculated, poorly marginated soft­ fine, very straight spiculations of a breast
tissue densities, the clinical history, physical carcinoma that extend directly to the skin
examination, and comparison with previous and cause retraction. 94
studies are necessary for appropriate man­ The evolving scar in the breast contracts
agement. On physical examination, scarring and shrinks as it matures in the first year or
uncomplicated by fat necrosis is perceived as two. The period of change for an individual
induration rather than a mass. patient is variable and does not depend on
Radiolucencies within the central area of the breast parenchymal type. The size of
soft-tissue density suggest scarring (Fig. 7). resection, volume of postsurgical fluid collec­
This mammographic feature is one of the few tion , and whether it was drained postsurgi­
found to be helpful by Mitnick. Fo The radio­ cally may affect the rate of scar formation. A
lucencies represent fat trapped by fibrous scar may have formed within 6 months or as
stranding in the parenchymal scar. Other late as a year or more after tumor excision
authors have noted a similar appearance with and breast irradiation. 56 On sequential stud­
lack of a central mass in postsurgical scar­ ies, decrease in scar size may be barely per­
ring, fat necrosis, and in nonencapsulated ceptible or seen in only one projection . After
sclerosing lesions (radial scars) as opposed two successive studies have shown no
to most breast carcinomas. Some carcinomas, change, recurrent tumor should be suspected
notably infiltrating lobular, may contain ra­ if there is increasing size or nodularity at or
diolucencies and may not have a central near the scar. Fine-needle aspiration cytol­
tumor focus 57. 58 ogy, core biopsy, or surgical excision may
Another finding observed in intramam­ confirm recurrent tumor as the cause of in­
mary scars is a changing appearance in dif­ creasing soft-tissue density. A diagnosis of
ferent projections. In one view, the spicu­ fibrosis , fat necrosis, or granuloma may be
lated soft-tissue densities are masslike, but more certain with surgical biopsy.
they elongate in other projections. More dif­
ficult discriminators to apply, in our experi­ Mammographic Evaluation of
ence, are the length, thickness, density, and the Scar
communication of spicules with the overlying
skin (Fig. 8). Wolfe describes spicules of Because breast density and architectural
scarring as curvilinear and thicker than the distortion cause interpretive uncertainty, the
116 MENDELSON

Figure 7. Mature scar 4 years after breast conservation therapy. A, Tangential view shows a fibrous bridge connecting
the parenchymal scar (large arrow) with incision site (short arrow) marked by thin wire taped to the skin. B, Another
projection showing the amorphous scar with thin spiculations and benign calcification . C, A third prOjection showing still
another shape of the scar and demonstrating radiolucencies of fat trapped by fibrous tissue with lack of a central mass.
Spot compression and magnification are essential in demonstrating features of scarring and excluding recurrent tumor.
 EVALUATION OF THE POSTOPERATIVE BREAST 117

Figure 8. Scarring versus carcinoma: Nonspecificity of spicule length. A, Magnification view of a tubular carcinoma
shows central radiolucencies and microcalcifications. Radial scar was also considered preoperatively. Long, thin spicules
extend toward the skin, of normal thickness. Radiopaque marker corresponds to a thickening felt by the patient. B,
Postoperative pre radiation study shows patch of soft·tissue density at excision site with thickened skin and very long
spicules extending toward the incision . Length, curvature, and thickness of the spicules represent only scarring here
but are difficult features to apply in differentiating malignancy from scarring.

standard mediolateral oblique and craniocau­ Increased Breast Density:

dal views require supplementation by addi­ Breast Edema and Skin Thickening
tional mammographic projections that dem­ Increased breast density caused by edema
onstrate characteristics of scars. The area of and skin thickening is one of the most strik­
excision should be imaged fully in two pro­ ing changes in the mammographic appear­
jections. Depending upon the location of the ance after lumpectomy and breast irradiation .
tumor, various obliquities can provide this
visualization. If the surgical site is near the Breast Edema
posterior edge of the compression plate, a A benign biopsy will cause breast edema
small spot compression device will be useful in the first month or two, but the mammo­
both in fixing the area and in spreading apart graphic changes are less marked and resolve
the tissue elements (Fig. 9). Magnification more rapidly than when the breast has been
radiography of the surgical site should be irradiated. After benign biopsies and lum­
performed in the projection in which it is pectomies for carcinoma prior to radiation
seen most completely. therapy, edema is often mild.
The thickened skin at the incision may be ~early all patients who have had axillary
superimposed upon the surgical bed causing dissection or radiation therapy have breast
masslike increased density, particularly edema. 8 • 37 :\1oderately severe or marked
when a keloid has formed. Csing a wire breast edema may be present on the mam­
taped to the scar as a focus for positioning mogram requested for the 6-month follow­
the patient, a view tangential to the scar up evaluation after radiation therapy has
permits separation of the skin and parenchy­ commenced. Thickened, stringy linear paren­
mal elements. 45 As discussed, the distinction chvmal trabeculations are seen when edema
between scarring of the skin and the surgical is ~oderately severe. The breast enlarges and
bed can be made sonographically as well. At mammographic compression is more difficult
the incision, a band of thickened skin will be as edema involves th e parenchyma, subcu­
seen. taneous tissues , and skin. In marked edema,
118 ME:-.JDELSUN

Figure 9. Use of spot compression in postoperative mammography. A. Craniocaudal view shows portion of a spiculated
nodular soft-tissue density in outer posterior breast where tumor excision (wire taped to skin) had been performed 3
years earlier. It is impossible to exclude recurrent tumor. B, Spot compression film separates the parenchymal elements
more completely, is able to engage the posterior tissue better than the larger plates , and demonstrates radiolucencies
within the scar (arrows). Recurrent tumor was excluded with confidence.

the breast is homogene ously dense and white Congestive heart failure and recurrent carci­
(Fig. 10). Outlined by subcutaneous fat, lin­ noma require exclusion. Although a new
ear densities representing engorged lym­ mass or microcalcifications are more common
phatics and interstitial fluid collections ex­ presentations of recurrent tumor, the reap­
tend toward the skin, which is thickened. pearance of breast edema can represent lym­
These changes are most evident in the peri­ phatic dissemination of breast carcinoma,
areolar and dependent areas of the breast. ;' which requires a different therapeutic man­
Breast edema gradually diminishes and re­ agement.
solves for many patients within 2 years. Mild As edema recedes, the breast size also
edema persists in a small percentage of pa­ normalizes. After several years, the irradi­
tients and is seen radiographically as coars­ ated breast may become progressively
ened linear in terstitial elements. IS. S2 In one smaller and somew hat denser because of
study, 20 % of women had edema lasting radiation fibrosis 6 l
more than 3 years, slowly resolving between
4 and 8 years after irradiation 37 The breast's Skin Thickening
lymphatic drainage is toward and through
the axilla, and a cause of prolonged edema Skin thickening and breast edema are com­
may be the interruption of lymphatic drain­ panion findings that have similar time
age in the axilla after extensive axillary dis­ courses for maximal change and resolution
section. after lumpectomy and breast irradiation. The
After breast density has stabilized, recur­ skin of the breast is ordinarily less than 0.2
rent edema is cause for concern. A small cm thick, but it is slightly thicker in the lower
percentage of patients will develop infection breast near the inframammary fold and
postoperatively and after radiation therapy. around the nipple. After radiation therapy,
 EVALUATION OF THE POSTOPERATIVE BREAST 119

Figure 10. Breast edema and skin thickening. A, Six months after lumpectomy and radiation therapy, craniocaudal view
shows marked edema of the breast parenchyma and skin , particularly in the anterior half and periareolar region. Tumor
excision site in the medial breast is marked by a wire taped to the skin. B, One year after breast conservation therapy,
moderate edema remains with stringy interstitial densities and periareolar skin thickening. Although edema still obscures
visualization of the fibroglandular tissue in this predominantly fatty breast, the lumpectomy site is better visualized.
C, Craniocaudal view of the same patient 1 year later (2 years after initial treatment) shows progressive contraction and
spiculation of the scar and continued resolution of edema. Mild-to-moderate edema persists, and the periareolar and
medial skin thickening is unchanged or minimally diminished (arrow).
120 MENDELSON

the skin's thickness may reach 1 cm or even dermis will be more echogenic, which reflects
greater (Fig. lOA) in the periareolar regi on. ' 6 fluid accumulation and fibrosis. Because so­
As in parenchymal edema, skin edema clears nography is recommended for evaluation of
last in the periareolar area and dependent postoperative masses and fluid collections, it
portion of the breast. By 2 to 3 years after is important that the sonographic appearance
breast conservation therapy, the skin has of the entire region including the skin and
returned to near-normal thickness, although tracts from the surgical bed to the skin's scar
mild thickening persists in approximately be recognized as well.
30% of the patients .52

Calcifications
Imaging Evaluation of
Increased Breast Density
Calcifications are a most important marker
Increased breast density, a reticular pat­ of new or recurrent breast carcinoma . Forty­
tern, and measurable thickness of the skin three percent of mammographically detected
are easily recognized mammographic signs recurrences were manifested by calcifica­
of skin thickening and edema, particularly if tions Y In evaluating calcifications at the lum­
only one breast is affected. Csing a bright pectomy site, radiologists should apply the
light to illuminate the skin line on high con­ same morphologic and di stributional features
trast film-screen studies will ensure that a u sed preoperatively to rate the probability of
focal area of skin thickening or subcutaneous malignancy . Postoperative preradiation mag­
abnormality will be noticed. Breast edema nification radiography should be performed
and skin thickening also can be assessed with to detect residual calcifications at the tumor
ultra sonography (Fig. 11). An offset pad or excision site (Fig. 12)73 The surgical bed or­
built-in standoff should be used to bring the dinarily will be reexcised if microcalcifications
skin and superficial tissues into the trans­ remain. ,t . 80 Mammographic localization may
ducer's focal zone. be necessary, and the specimen should be
The 0.2-cm-thick skin complex is com­ rad iographed. The surgical procedure is eas­
prised of two thin echogenic lines with a ier to accomplish, with fewer sequelae, if it
hypoechoic dermis between them . 56 As in ter­ is performed before breast irradiation .
stitial fluid seeps through the tissues during It is common for new calcifications to occur
the first year after radiation therapy, the at the site of tumor excision in conservatively
deeper of the two lines becomes interrupted. treated breast cancer patients. ~ew calcifica­
Curved arcs are seen extending through the tions w ere found within 6 to 12 months in
subcutaneous fat to the thickened skin on 28 o/c of 110 patients treated with breast con­
the mammogram, and interstitial fluid is im­ servation therapy whom we studied. 56 Al­
aged sonographically as linear collections in though some calcifications appeared within
nonductal distribution extending toward the 6 months, in many patients additional calci­
skin. 56 As edema diminishes, the deeper line fications were seen over a period of 3 to 4
resumes continuity. After radiation therapy years. In unpublished data , Paulus found
or with any process that causes edema, the that 20% to 25 % of 450 irradiated breasts

Figure 11. Skin thickening : so­

nography . A, Six months af1er tu­
mor excision and irradiation, mag­
nified sonogram (7 .5 MHz with
offset pad) demonstrates skin
thickening to 1 cm. B, Follow-up
sonogram 6 months later shows
decrease in skin thickening . Skin
measures 0.5 cm thick . Cursors
denote the superficial and deep
echogenic lines of the skin com­
plex. Between them is the dermis,
more echogenic than seen nor­
mally, suggestive of edema or fi­
brosis.
 EVALGATIO N OF THE POSTOPERATIVE BREAST 121

fine, faint, and difficult to characterize as

they begin to precipitate in scars, necrotic
tissue, or areas of fat necrosis. Most studies
assign a low probability of malignancy to
calcifications that occur soon (6-18 months)
after the surgery and radiation therapy.4o In
Rebner's small patient group, there was con­
siderable overlap, and new malignant micro­
calcifications antedated benign calcifications
at the lumpectomy site Y Although the time
course may not always be reliable, it serves
as a reasonable guide to interpretation of
alterations in mammographic appearance,
with early changes often being benign.
Fat necrosis is associated with all types of
surgical procedures in the breast. In vague,
pa tchy soft-tissue densities, calcifica tions
may be needlelike, of varied shapes, bizarre,
disorganized appearing, and alarming. These
dystrophic calcifications will become thick,
calcified plaques. Common expressions of fat
necrosis are thin arcs of calcification, which
may form complete circles to define the rims
of rounded , radiolucent oil cysts 6 Oil cysts,
fibrous tissue, and calcifications may form
calcifying conglomerations of tissue present­
ing as palpable masses on physical exami­
nation. Radiographically, the radiolucent
centers of these heterogeneous masses will
suggest their benign postoperative etiologies
(Fig. 13) thus leading to a recommendation
for follow-up evaluation rather than biopsy .
The physical findings, however, may be wor­
Figure 12. Residual carcinoma. A, Compressed, magni­ risome enough to prompt rebiopsy.
fied view of specimen . Near the tip of the localization Developing unilaterally after tumor exci­
hookwire is a stiffening cannula placed over it at surgery
to facilitate its palpation intraoperatively. Microcalcifica­ sion and radiation therapy, small , round , and
tions are at the margin (arrow) of the resected tissue. smooth dystrophic calcifications are common
After comparing the specimen radiograph with the mam­ at the lumpectomy site. They resemble the
mogram, the surgeon, still in the operating room, was calcifications of secretorv disease or ductal
advised to remove additional tissue, but no further spec­
imens were received for radiography. B, Postoperative
ectasia and may have a similar pathogenesis,
preradiation magnification view of the scar in mediolateral forming in areas of necrotic tissue, sloughed
oblique projection shows wire taped to the linear scar. cells, and cellular detritus. 57 Paulus found
More posteriorly, an area of vague soft-tissue density with calcifications of this type in 20 % to 25 0/, of
microcalcifications (arrow) represented residual comedo­ mammograms of conservatively treated
carcinoma in the surgical bed. The microcalcifications
were relocalized and removed. breast cancer patients that he reviewed 65 As
new unila teral calcifica hons in the trea ted
breast, a causal relationship is inferred be­
tween the procedures and the occurrence of
developed benign appearing calcifications .6' the calcifications.
Libshitz, studying a smaller group of patients Also benign appearing and somewhat less
undergoing tumor excision and radiation common are more coarse, plaquelike, angular
therapy, noted that 14 0/, of 81 patients de­ calcifications. These larger calcifications are
veloped benign calcifications during a wide­ also dystrophic, developing in scars and in
ranging period of 2 to 44 months. 51 the disturbed subcutaneous tissue beneath
Several types of benign calcifications are the incision. Views tangential to the skin can
found at the lumpectomy site. Although they demonstrate their superficial locations.
will coarsen later, calcifications may be quite Calcified remnants of suture material at
122 ME:',JOELSON

Figure 13. Calcifications at the

tumor excision site: fat necrosis .
Three years after tumor excision
and irradiation , the parenchymal
scar contains radiolucencies, and
curved spicules extend to the
thickened skin of the incision (ar­
row). A cluster of calcifying oil
cysts indicative of fat necrosis is
present. Other benign calcifica­
tions, dystrophic, are seen at the
site (short arrows) .

the lumpectomy site have distinctive shapes series. 87 Mammographic analysis is made
(Fig. 14). Knots, branching linear calcifica­ more difficult by the increased soft-tissue
tions, and double tracking may suggest ma­ density and architectural distortion of the
lignancy, but these linear calcifications can parenchymal scar and possible confusion
be several millimeters long and quite wide. with benign calcifications at the excision site.
They resemble the thick linear calcifications At the outset, benign forms of calcification
of ductal ectasia or secretory disease. 45 . ,; . 95 should be excluded and an active search,
Davis suggests that magnification radiogra­ with magnification radiography, made for
phy of the lumpectomy site will remove any new indeterminate or malignant-appearing
hesitancy in calling them benign. :9 They do calcifications (Fig. 15). Rebner cautions that
not require biopsy. Although sutural calcifi­ there is overlap in appearances and recom­
cations are relatively rare, encountered in the mends that unless the calcifications are un­
mammograms of 3 of 110 patients we stud­ equivocally benign, they be biopsied Y
ied, once having been recognized, they will Twenty-seven (18%) of 152 patients devel­
subsequently be an "Aunt Minnie." oped calcifications. Of those, 10 patients had
Microcalcifications are the most common malignant or indeterminate calcifications and
radiographic sign of recurrent tumor, with had biopsies (7%.). Four biopsies of microcal­
43 % of mammographically detected recurr­ cifications were positive for malignancy in a
ences presenting in this fashion in Stomper's period of 6 to 32 months. In 75%, tumor had

Figure 14. Sutural calcifications.

Magnification view of lumpectomy
site 4 years after tumor excision
and radiation therapy for infiltrat­
ing ductal carcinoma shows
coarse calcifications, unchanged
from their appearance 1 year ear­
lier. Wishbone-shaped strands
joined by a knot (arrow) are seen
at one end of the thin wire marking
the scar. Another linear calcifica­
tion (curved arrow) with a knot in
its midportion also suggests
clipped suture material. Additional
benign calcifications may repre­
sent suture material , other dystro­
phic calcification, or fat necrosis.
 EVALLATIOl\ OF THE POSTOPERATIVE BREAST 123

Figure 15. Tumor recurrence : microcalcifications. A. Five years after lumpectomy and radiation therapy for infiltrating
ductal carcinoma, no mass is identified at the lumpectomy site. On this unmagnified view, faint new microcalcifications
were seen at the posterior aspect of the scar (arrow). B, Lateral spot compression magnification view of the tumor
excision site shows architectural distortion and a linear soft-tissue density containing numerous, pleomorphic microcal­
cifications (arrow) that represent a new focus of ductal carcinoma in situ.

presen ted ini tially with microcalcifica tions. diation therapy, and their persistence does
Although Rebner's numbers are small, they not necessarily indicate viable tumor. 51 The
emphasize the need for a high degree of expectation of recurrence will be higher after
vigilance in follow-up studies of the lumpec­ excision of invasive carcinoma with extensive
tomy site, particularly when the tumor has intraductal carcinoma or with large areas of
presented as microcalcifications. comedocarcinoma 37 Accuracy of interpreta­
Another study makes a similar pOint 8 1 tion will be furthered by familiarity with the
With an overall positive biopsy rate of 52% clinical background, careful review of se­
for recurrent tumor (in the loca tion of the quential studies, and magnification radiog­
primary or in other quadrants), 21 G/o of 145 raphy of the lumpectomy site and other areas
biopsies were performed for microcalcifica­ where calcifications are suspected. 73
tions seen on the mammogram without as­
sociated physical findings. Of this group, RECURRENT TUMOR:
66 % were positive, a much higher percentage EFFECTIVENESS OF
after breast conservation than for patients IMAGING
recommended for biopsy of microcalcifica­
tions after routine screening mammography. Berenberg states that edema, fibrosis, and
In summary, most newly occurring calcifi­ architectural changes may make mammog­
cations in the treated breast are benign. Re­ raphy less conclusive after breast conserva­
sidual microcalcifications should be excluded tion therapy, but he notes that in six local
prior to radiation therapy. C'nexcised calcifi­ recurrences found in 126 cases (4.8 %), the
cations mayor may not disappear after ra­ mammogram alone was positive in four of
124 MENDELSON

the six. 7 The other two recurrences were tified 48 months after lumpectomy and radia­
detected through physical findings, not tion therapy were not imaged, even with
mammography. The data of Stomper are sim­ knowledge of their precise location, by ultra­
ilar: 35% of 45 recurrent carcinomas after sonography, One suspicious palpable mass
lumpectomy and radiation therapy were de­ not detected mammographically was seen
tected only by mammography and 61 (I, of with ultrasonography, Several palpable
tumor recurrences were identified through masses with a sonographic appearance con­
mammography with the addition of clinical sistent with either fat necrosis and scarring
findings Y Berenberg's, Stomper's, and other or recurrent tumor and no mammographic
studies 34 ,40 support the conclusion that mam­ change represented fat necrosis on biopsy.
mography is an effective technique for fol­ Without specifying the presentation of tu­
low-up evaluation of these patients but that mor recurrences, Balu-Maestro and her
mammography requires complementation group in France found that mammography
with physical examination to maximize de­ and sonography allowed high detection rates
tection of recurrent tumor. Similar results of tumor recurrences. 4 Ninety-five and one
and conclusions are offered by Fowble, half percent of tumor recurrences were iden­
whose group reported on 66 tumor recur­ tified mammographically and 90,9% by ultra­
rences after breast conserva hon therapy in sonography, Only 45,5 % were detected by
which there was mammographic detection phYSical examination, 4 In no instance did
exclusively in 29% (13119 presenting with sonography alter the management plan for
biopsy or fo!Iow -up study that was deter­
microcalcifications without a mass) and phys­
mined by physical or mammographic find­
ical examination alone in 50% (11165 did not
ings,
have mammography, however, at the time
Nevertheless, sonography is an excellent
the recurrent tumor was diagnosed) and with supplemental technique, and familiarity with
both in 21 %.32 In all of these studies, micro­
the sonographic appearances of findings after
calcifications were found to be a major sign tumor excision and radiation therapy is nec­
of recurrent tumor that cannot be identified essary to avoid misinterpretation, In the post­
by physical examination and requires mam­ operative breast, sonography can be used to
mography, characterize a mammographiC or palpable
Of the imaging techniques, mammography mass as fluid-filled or solid, to guide inter­
has been studied the longest, and its limita­ ventional procedures, and as an alternate
tions in the densely fibroglandular, early method of imaging sequen tial changes such
postoperative, edematous breast are ac­ as skin thickening.
cepted, Second in frequency of use, with Gadolinium-enhanced MR imaging is un­
specific applications to the management of der investigation for distinguishing mature
breast disease, is ultrasonography, Microcal­ scarring from recurrent tumor at the lumpec­
cifications are the most frequent indicator of tomy site. 4}, 4-1 Theoretically, a seasoned area
recurrent tumor, and ultrasonography dem­ of postsurgical change would not show gad­
onstrates microcalcifications poorly 20,28, 47 So­ olinium uptake , but an active area of tumor
nography cannot be substituted for primary growth might. Sensitivity, specificity, and
evaluation of the breast before or after sur­ possible indications for this technique have
gery, not yet been established,
In our practice, ultrasonography has been For postprocedural findings at the lumpec­
useful in the fo!Iow-up evaluation of the tomy site, pulsed Doppler and the more
conservatively treated breast cancer patient. sensitive color flow imaging have proved
In serial mammographic 6-month follow-up disappointing, adding no useful diagnostic
studies of 110 patients for at least 3 years, information apart from that provided by
postoperative masses in the first year were mammography and standard sonography.11
identified unequivocally as fluid collections Cosgrove, investigating usefulness of color
in 40.5% of patients, 56 Recurrent tumor at or flow Doppler in breast masses, found no
within 2 cm of the lumpectomy si te was abnormal vascularity in five of seven locally
subsequently detected in 6% of these treated recurrent breast malignancies. 18
breasts, the earliest at 30 months, Three
masses without calcifications and one with PROTOCOLS FOR FOLLOW­
calcifications were detected with mammog­ UP IMAGING AFTER BREAST
CONSERVATION THERAPY
raphy, with the masses also being demon­
strable sonographica!Iy and distinct from the There is considerable geographic variation
scars (Fig, 16), ;'\Jew microcalcifications iden­ within the enited States in the treatment of
 EVALUA nON OF THE POSTOPERATIVE BREAST 125

Figure 16. Recurrent tumor: mass . A, Two years after breast conservation therapy , minimal breast edema is present.
The lumpectomy site in the outer aspect of the breast on this craniocaudal projection appears linear and spiculated . B,
One year later, a new mass (a rrow) is present posteromedial to the scar, representing recurrent infiltrating ductal
carcinoma.

breast cancer. Tumors of the same stage and recurrence (high positive predictive value
size may be trea ted altogether differently when rebiopsy is suggested). Achievement
depending on the philosophy in a given of these objectives will be facilitated by fa­
locale. On the two coasts, the majority of miliarity with timing of tumor recurrence and
breast cancer patients may be treated with the sequential postoperative and irradiation
wide excision and irradiation, but in some changes that are anticipated: masses and
areas up to two thirds of women eligible for fluid collections, scarring and architectural
breast conservation are not being offered this distortion, edema and skin thickening, and
option. 72 Somewhat dependent on the pre­ calcifications. In Table L results of our on­
vailing surgical practices, there is also wide going study of these findings are summa­
variation in experience with th e radiologic rized.
follow-up evaluation of these patients . Cur­ The treated breas t is a rapidly changing
rently, beyond the first year after therapy, organ, with changes as dramatic as those
no guidelines have been adopted for intervals seen at puberty. Followi ng lumpectomy and
between follow-up studies, although many radiation therapy, the mammogram will por­
recommendations have appeared in the lit­ tray the magnitude of the assault required to
erature .40. 64. 65. 75 control the disease process . Later, the mam­
To determine an appropriate schedule of mogram will depict the resilience of mam­
studies, the purposes of imaging the conser­ mary tissue as the breast returns to a more
vatively treated breast must be clearly stated . normal appearance. On the basis of radio­
Two major objectives are (1) early dia gnosis graphic evolution (progression) or resolution
of recurrence, prior to development of me­ of changes, several periods are defined (Fig.
tasta ses,81 and (2) minimizing misin ter­ 17) .
pretations of postprocedural change as tumor The period of greatest change occupies the
126 YlENDELSO N

first 18 months, with the most marked notes that patients whose recurrences were
changes occurring around 6 months. 40 If the diagnosed mammographically fared better
margins of resection have been free of tumor than those who presented clinically, and she
and the breast has been irradiated, recurrent argues for close interval follow-up evalua­
tumor is unlikely to explain a large masslike tion, Her Vancouver group advocates mam­
density on the mammogram (Fig. 18). mography at intervals of 6 months for 3
Perhaps the most important period is sta­ years, then annually ,
bilization of the breast, which we have de­ Our recommendations for imaging after
fined as lack of change on two successive breast conservation therapy are similar to
mammograms . Subsequently, a change Hassell' s and are shown in Table 2, L'ntil
counter to the direction of resolution would 1988, when we extended to 5 years the period
suggest recurrent tumor. In an individual in which we performed mammography at 6­
patient, selection of the appropriate follow­ month intervals, we requested mammograms
up interval for mammography depends on at 6-month intervals for 2 years with annual
determination of stability. Many proposals examinations · after that. We currently pro­
for the time of stabilization have been of­ pose studies every 6 months for 3 years to
fered. For Berenberg/ the conservatively cover the 1 to 3 year period of stabilization
treated breast is stable at 4 months, for as much to avo id unnecessary biopsies of
Stomper 6 at 3 to 6 months, and for Paulus,65 postprocedural benign changes as to diag­
6 to 8 months. The term stable is not defined, nose recurrent tumor.
and no criteria for establishing this milestone The achievement of stability coincides with
are specified. Differences might be explained
the time that recurrences begin to appear,
by variations in surgical procedures such as
which is 2 to 3 vears after conservation ther­
size of resection or drainage of the lumpec­
apy. A new m~ss, microcalcifications, or ar­
tomy site, the radiation dose , and in the
chitectural distortion at the lumpectomy sIte
definition stability .
There is agreement that mammography at may be more easily recognized as an interval
6 and 12 months after tumor excision will change when the breast is stable, and after 3
record the greatest changes in the postpro­ years, an annual study may be sufficient for
cedural breas t. Beyond th e I-year stud y, var­ most patients. Women treated with tumor
ious schedules have been proposed for fol­
low-up mammograms . Most authors have
Table 1. FOLLOW-UP AFTER BREAST
supported studies every 6 month s after the CONSERVATION THERAPY :
first year for some period of time 40. 56, 73, 74 MAMMOGRAPHIC FINDINGS
Sickles comments that he reevaluates his IN 110 PATIENTS
lumpectomy and radiation therapy patients
% of
every 6 months for a "few years," 7? Cady
Total Time
advocates 6-month follow-up studies for 2 Findings Patients Period
years.16 Rebner calls for mammography at 6­
Calcifications 52 75% by
month intervals for the first 2 years, then 18 mo,
annuall y for the treated breast and the con­ Mass or scarring 100 by 2-3 Y
tralateral breast, which is also at increased Postoperative fluid collections 40 at 6 mo
risk of carcinoma,~5 , ; 4, 67 The recommenda­ (US confirmation)
Spiculated densities and 60 at 6 mo
tions of Paulus are for mammograms to be architectural distortion
performed after the initial 6-month study Breast edema (mild, moderate, 100 6-12 mo
either annually or at 6- to 9-month intervals marked)
for several years. 64 , 65 Hassell, reporting on moderate to marked 72 6-12 mo
moderate to mild gradual 12-18 mo
rebiopsies of 48 conservatively treated breast
decrease
cancer patients, sugges ts that more frequent resolution 60 2Y
follow-up evaluation in the first several years Skin thicken ing
might help reduce false-positive interpreta­ range: 0.4-1,0 cm (average) 100 6-12 mo
50% reduction 50 18-24 mo
tions of recurrent tumor at the lumpectomy Recurrent tumor 5,5 30-48 mo
site, particularly during the early period Masses detected 3
when biopsies of thi s area are most often mammographically
benign . 40 Although she states that the num­ New calcifications
Calcifications and mass
ber of recurrences in her continuing study is
Palpable mass
too small to suggest general guidelines, she
 EVALLATlON OF THE PUSTOPERATIVE BREAST 127

CHANGING APPEARANCE OF THE CONSERVATIVELY-TREATED BREAST

~·-·-·-·-·-s·t~bili;;ti;~-·-·-·-·-·-· •••••••••• Increasing Frequency
...........................
: Most Rapid Change •
of Recurrence
I·········~··········.
: Most Mar~d Change :
100~·~~~·----~:~----~-------+----~~==========~
Scarring/Fibrosis

Calcifications

Edema

36 mo. 42 mo. 48 mo.

Figure 17. Changing appearance of the conservatively treated breast.

excision and irradiation who have a known In general, the breast tissue and pectoralis
increased risk of recurrence, as in infiltrating major muscle are removed with a generous
carcinoma with an extensive intraductal com­ ellipse of skin overlying the tumor. The area
ponent,37 might benefit from increased mam­ of incision extends to the axilla, which is
mographic vigilance through continuation of completely or partially dissected, Modified
the studies every 6 months of the treated radical mastectomy offers disease-free sur­
breast. vival equivalent to that of radical mastectomy
with better cosmetic result. In addition, re­
construction is easier to accomplish if the
IMAGING AFTER pectoralis major muscle is preserved.
MASTECTOMY Mastectomy should leave no breast tissue
for mammographic evaluation, and ipsilat­
Until the 1970s, the radical mastectomy eral axillary dissection removes nodal sites of
was the surgical procedure of choice for treat­ potential spread, Recurrent local breast car­
ment of breast cancer in the United States. cinoma in more than 50 % of cases involves
Removed en bloc are the breast, skin over the chest wall or overlying skin. 70 Recurrent
the tumor, both the pectoralis major and tumor is frequently detected in physical ex­
minor muscles, and the axillary contents. J9. 49 amination, and CT can be used to evaluate
Although the radical mastectomy remains an its extent.
option for patients with bulky tumors involv­ Should the axilla on the side of mastectomy
ing the pectoralis major muscle or fascia, the be imaged? No studies in the literature sup­
modified radical mastectomy has become the port its use in the patient who has had an
more common surgical procedure. 1s ,;9 Pa­ axillary dissection, Wolfe suggests a lateral
tients with Stage I to III tumors not fixed to view of the axilla for breast cancer patients
the pectoralis major muscle by axillary lymph who have not undergone axillary lymphad­
nodes are candidates for modified radical enectomy or for patients with complaints
mastectomy, which has a number of variants, referable to the axilla 93
128 MEN DELSON

Figure 18. Mass at the site of tumor excision. A, Preoperative magnification view shows a spiculated mass with
microcalcifications. B, The infiltrating ductal carcinoma was resected and radiation therapy performed . The site of
excision is marked with a wire taped to the skin . One year later, a large oval soft-tissue density is seen with some
stranding in its posterosuperior margin. Recurrent tumor of this size is unlikely, especially in the first year after treatment.
C, Sonogram shows a well-defined hypoechoic mass with septa and locules, a solidified postoperative hematoma that
will be very slow to resolve . Mass was firm , and there were no signs of infection.
 EVALUATIO N OF THE POSroPERATIVE BREAST 129

Table 2. IMAGING AFTER BREAST have masses, pain, or other complaints refer­
CONSERVATION able to the axilla, the view can be retained.
THERAPY In some instances, a mammogram will be
Timing of the Study reassuring to the patient in showing that
Treated Breast Rationale asymmetric lumpiness at the operative site is
Postoperative preradiation Detection of residual fatty redundant tissue.
(2-5 w after carcinoma Mammographic study of the contralateral
lumpectomy) breast can be performed according to screen­
6 mo Baseline study ing guidelines or as indicated by clinical and
Peak of postprocedural
changes: masses. skin
radiographic considerations. In assessing the
thickening. edema. remaining breast, the radiologist should be
Early calcifications aware of the histology of the carcinoma, its
form. incidence of bilaterality, and presence of ad­
12 mo Assess changes listed ditional indicators of risk for breast carcinoma
above; begin to look for such as lobular carcinoma in situ (lobular
mammographic stability
(no change on 2
neoplasia) that may increase the likelihood
successive studies) of malignancy.
18 mo End of time of most rapid
change; confirm stability IMAGING AFTER COSMETIC
24 mo Expect stabilization for SURGERY
most patients. More
confident recognition of Breast Reconstruction
benign postprocedural
changes . The diagnosis of breast cancer inflicts psy­
30 mo Mammogram should be chologic trauma. Breast reconstruction may
stable for nearly all provide some restoration of body self-image
patients. after mastectomy. Pros and cons of breast
36 mo Stable mammogram.
Suspect recurrence if
reconstruction and severa l methods of ac­
direction of change is complishing it are presented to the patient
unexpected. as therapy is discussed. Breas t reconstruction
Annually Detect recurrence. should meet reasonable cosmetic expecta­
For patients at increased tions without compromising oncologic con­
risk of recurrence (EIC trol and management. 1;
and young patients),
consider intervals of 6
Reconstructive surgery can be performed
months. immediately, months, or years after the mas­
Contralateral breast tectomy. There are two types of breast recon­
Annually Screening; increased risk struction procedures: implants or autogenous
of breast carcinoma tissue transfer (myocutaneous flaps or free
flaps attached with microvascular technique).
At times a combination of both may be used
The lateral view of the axilla is obtained with reduction, augmentation, or mastopexy
without compression, with the molybdenum of the contralateral breast to achieve sym­
filter of the mammographic unit replaced by metry.
one of aluminum, using a higher kilovoltage Silicone implant reconstruction can be per­
and milliamperage than are used to image formed at the time of mastectomy or later.
the breast. Submuscular placement is preferred to min­
In our practice, for the last 3 years, we imize complications such as fibrous contrac­
have not imaged the axilla routinely on the ture, which commonly occur after placement
side of mastectomy and axillary dissection. of silicone prostheses, particularly if the im­
Prior to that, between 1985 and 1988, the plant is anterior to the pectoralis major mus­
only abnormality seen was bone metastases cle. Retropectorallocation also compromises
already identified in a symptomatic patient. mammographic and physical examination of
In personal communications with colleagues, the reconstructed breast less than a sub­
many agree that the lateral axillary view is glandular installation.
noncontributory and have eliminated it from The mastectomy incision is frequently
their routine imaging of patients with mas­ used, and the implant is inserted into a
tectomies who are asymptomatic or have pocket located behind the pectoralis major.
unchanged arm edema. For women who When the space is too small to accommodate
130 ME:"-JDELSOf\J

the implant, a saline-filled tissue expander is nal wound is then closed. Hernias involving
used to enlarge the space. Creation of a the donor site are not uncommon.
pocket large enough for a prosthesis may Because tumor frequently involves the nip­
require multiple outpatient visits for injection ple and areolar tissue, the nipple-areolar
of increasing amounts of saline. Two or three complex is not preserved at the time of mas­
surgical procedures also will be required to tectomy for use in reconstruction. Donor sites
complete the prosthetic reconstruction. for the areola include postauricular skin and
Autogenous tissue transfer is more com­ that of the upper inner thigh. Tattooing can
plicated technically than placement of a pros­ be used to match the skin color to that of the
thesis. It involves more extensive surgery but contralateral areola. A nipple can be created
offers more natural-looking breast simula­ by a variety of flap techniques or transplan­
tion. Myocutaneous flaps and microvascular tation of a portion of the contralateral nipple.
free flaps transport distant tissues to the The necessity to image the reconstructed
mastectomy site for creation of a replacement breast has not been established. The mastec­
breast mound. The latissimus dorsi flap tomy patient with and without a simulated
swings anteriorly on its vascular pedicle to breast shares the same risk of tumor recurr­
the mastectomy site. With this flap, an im­ ence. Although film-screen compression
plant is often necessary to achieve the appro­ mammography can be performed on a recon­
priate bulk of tissue. structed breast either with an implant or with
The transverse rectus abdominis myocuta­ a myocutaneous flap, in asymptomatic pa­
neous (TRAM) flap may achieve the best tients with myocutaneous flap reconstruction
cosmetic result and does not require a pros­ as with the unreconstructed mastectomy pa­
thesis. TRAM flap reconstruction is major tient, because of the low yield of significant
surgery, doubling the time of mastectomy if radiographic findings, we no longer perform
reconstruction follows immediately14. 92 In routine follow-up studies. If findings on
this procedure, transverse abdominal inci­ physical examination or breast self-examina­
sions are made, forming a large ellipse of tion suggest an abnormality, however, mam­
subcutaneous abdominal tissue and skin (Fig. mography will be performed.
19). On a vascularized pedicle, this flap, With mammography, no radial ductal or­
attached to the rectus abdominis muscle, is ganization is seen. Behind the created nipple,
tunneled into the mastectomy site where it no ducts are present. The simulated breast is
is shaped and sewn into place. The abdomi- fatty with vascular and connective tissue ele-

INrKAMA~~MA

RECTUS
ABDOMINIS M. Figure 19. Transverses rectus abdominis my­
ocutaneous (TRAM) flap reconstruction. An
elliptic incision is made isolating a tissue flap
including skin and subcutaneous fat. Maintain­
ing its communication with the rectus abdom­
SKIN
inis muscle, the flap is brought through the
ISLAND---~.oE;..""";:~""""~~ mastectomy incision into the site of breast
removal, and the breast mound is fashioned.
The abdominal incision is repaired, and the
umbilicus restored. (Adapted from Krizek TJ:
Breast reconstruction after mastectomy. In
Harris JR, Helman S, Henderson et al: Breast
Disease, ed 2. Philadelphia, JB Lippincott,
1991, P 496; with permission.)
 EVALLATIOl\ OF THE POSTO PERATIVE BREAST 131

ments scattered randomly (Fig. 20). Clips women elect the procedure annually,26 nearly
may be seen near the muscular pedicle. The the number expected to receive the diagnosis
skin at the sites of suturing may be thick­ of breast cancer each year.
ened. Injections of liquid silicone into the breast,
Most often, in myocutaneous breas t recon­ which are now illegal in the United States,
structions, a palpable mass may be explained have caused formation of calcified silicone
by fat necrosis.14 Patchy areas of soft-tissue granulomas, masses, chronic sinus tracts,
density are seen initially. Fine, faint calcifi­ and drainage .13 Elastomer-enclosed single lu ­
cations may develop, which become thicker men silicone implants and the saline injecta­
and heavier in time. Conglomerations of oil ble double lumen (outer lumen contains sa­
cysts, calcified and uncalcified, may occur. line and innerlumen silicone gel) prostheses
Despite a confident radiologic diagnosis of are common among the presently used
fat necrosis, worrisome physical findings prostheses. 13. 42
may result in biopsy. Associated complications include fibrous
Nodules of recurrent tumor at incision sites and calcific contracture; postoperative infec­
will be evident clinically and chest wall re­ tion in rash; implant rupture with escape of
currences may be studied with CT . free silicone into the parenchyma, axillary
lymph nodes and ducts; deflation of saline
prostheses; and possible association with
Augmentation Mammoplasty scleroderma and other connective tissue dis­
orders .13. 26. 27
Indications for augmentation mammo­ The silicone prosthesis can be positioned
plasty are for breast reconstruction after mas­ via periareolar, axillary, or inframammary
tectomy, for achievement of breast symme­ incisions either anterior to or behind the
try, for correction of congenital deformities, pectoralis muscle. The incidence of capsular
and for improvement of self-image. More than contractures is higher with subglandular sil­
1.5 million women have undergone breast icone implants, occurring in up to 74% of
augmentation since the early 1960s when sili­ patients in one study.13 It is speculated that
cone gel implants were developed so. 78; 150,000 the overlying muscular thickness compresses

Figure 20. TRAM flap with painful mass 5

years after reconstruction . A, Craniocaudal
view shows rounded, spiculated soft-tissue
mass. B, Exaggerated craniocaudal projec­
tion allows the mass to be seen completely.
Central radiolucency suggests a large oil
cyst, and the surrounding calcifications
(curved arrows) are compatible with fat ne­
crosis as well. Surgical clips (arrow) are
noted in the muscular portion of the flap.
Most of the flap is fatty, and no ductal
structures are present. Patient had felt a
new mass, presumably fat necrosis, that
was aspirated by the surgeon and became
infected.
132 MENDELSOl\:

and massages the implant between the mus­ Eklund reported on modified compression
cle fibers and the chest wall, thereby inhib­ views with which he supplemented the stan­
iting development of spherical capsular con­ dard mammographic technique in over 250
tracture and minimizing the unnatural patients 26 For the additional views, in cran­
firmness of the prosthesis. 13 iocaudal and mediolateral projections, the
Imaging and physical examination of the implant is displaced posteriorly against the
breast augmented with a retropectoral im­ chest wall while the breast tissue is pulled
plant is less compromised than with a sub­ over and in front of the implant and com­
glandular prosthesis. 21 The position of the pressed fully for the image (see Fig. 21). The
implant is more secure with the retropectoral compression paddle keeps the implant from
location, but submuscular placement is more reentering the field . This technique allowed
difficult in the tight, hypertrophic pectoralis 2 to 5 cm of additional breast compression,
muscles of athletic women (Fig. 21). The and no ruptures were reported. Eklund notes
submusculofascial placement is preferred by that for 150/, to 20% of patients with fibrous
most plastic surgeons to the subglandular encapsulation of the implants, the technique
wherever possible.1 3 was more difficult to accomplish. For all
The possible reduction in detectability of patients with implants, the standard views
breast carcinoma is of chief concern with are supplemented by the modified views,
cosmetic augmentation. Silverstein con­ and for patients with fibrous encapsulation,
cluded in a study of 20 women with implants he has added a 90-degree lateral view to the
and breast carcinoma, 65% of whom were standard views for depicting the tissue above
node positive at the time of diagnosis, that and below the implant. The standard views
the prosthesis might have caused delay in are used to image the posterior breasts. These
diagnosis. 78 four views, two standard and two with mod-

Figure 21. Modified views for implants. A. Mediolateral oblique projections (MLO) of each breast obtained with routine
positioning show hypertrophic pectoralis muscles surrounding the silicone implants. Only a minimal amount of
fibroglandular tissue is seen. B, MLO projections of each breast after the implant has been displaced posteriorly and
the anterior breast tissue pulled into the compression plates demonstrate only a little more parenchyma. C, Craniocaudal
views with the implants displaced posteriorly demonstrate periareolar dermal calcifications that were not seen on the
routine projections. Use of the modified views for the anterior tissue along with standard projections for the posterior
breasts has partially compensated for limitations on mammography imposed by the radiopaque implants. The positioning
techniques are more effective in the retropectoral implant location than the subglandular.
 EVALUATION OF THE POSTOPERATIVE BREAST 133

may rupture with escape of free silicone into

the breast (Fig. 22). Focal herniations or di­
verticula of the envelope may occur.23 Partic­
ularly with special views, mammography can
often identify the cause of an implanted­
related palpable massY Sonography also can
frequently delineate a locule within the im­
plant that has simulated a parenchymal
breast mass (Fig. 23).56 Although in one case
we were accurate in determining the site of
rupture in an implant with sonography, cap­
sular wrinkles and folds often mimic breaks. 56

Figure 22. Ruptured implant. Silicone globules are seen

in the axilla (small arrows) to be taken up by lymph nodes
and possibly to form calcified granulomas. The patient
had presented with a large mass, a diverticulum-like
outpouching or contained rupture of the implant (large
arrow). The implant is seen posterior to the pectoralis
muscle (curved arrow).

ified positioning, have been adopted in many

practices as the protocol for mammography
of the augmented breast.
Leibman and Kruse reviewed 11 cases of
breast cancer in patients with implants. "oC
Sixty percent of the patients were node neg­
ative, and 10 of the 11 patients presented
with an abnormal mammogram or sonogram.
With use of special mammographic tech­
niques including modified compression
views and sonography, these authors sug­
gest that early detection of carcinoma should
not be an unattainable goal after breast aug­
mentation. Figure 23. Palpable mass in augmented patient. A, Sub­
A mass palpated in the augmented breast glandular silicone implants in routine projections show no
may represent an abnormality of the breast contour abnormality of the implants and no mass in the
parenchyma of the right breast. B, Sonography excludes
tissue itself or the implant. 23 The implant's parenchymal abnormality and shows that mass is a locule
contour may change, becoming spherical (arrows) within the implant. On this scan, the pectoral
with fibrous encapsulation. 21 The implant muscle (M) can be seen posterior to the implant.
134 MENDELSON

Calcifications develop in the breasts after

augmentation and all other surgical proce­
dures. Thin plaques of calcification or bulkier
accumu lations can form on the capsules . The
radiolucent centered coarse eggshell calcifi­
cations of silicone granu lomas can be seen
after implant rupture or with the injection of
free silicone . Dystophic calcifications and
those of fat necrosis may form. Indeterminate I
I ,'--', I
microcalcifications near the capsule may rep­ I " \ I
resent postoperative fibrosis but will require
biopsy.
\
\
\
:
\
n)
."
',~' I
,
:

Because a variety of calcifications can be \

\
-.-""
I I
I
, I I
expected to develop after augmentation
mammoplasty, we support Leibman and "
' ......... - :: ",,'" "
------
Kruse's suggestion that presurgical and post­
surgical mammograms be performed for
these patients to ease the task of distinguish­ Figure 24 . Reduction mammoplasty. A circumareolar
ing benign postoperative findings from more incision is made with a vertical incision connecting it to
worrisome pare nchymal changes. The four­ an inframammary incision. Keyhole flaps permit removal
view study has improved the quality of mam­ of excess breast tissue. At the conclusion of the proce­
dure. the nipple, which has remained attached to a
mographic examination of these patients, vascu lar pedicle, is sewn into place in a higher location .
and the modified techniques of positioning
also can be applied to procedures such as
presurgical needle hookwire localizations. Radiographic features of reduction mam­
Use of the modified views minimizes chances moplasty include redistribution of tissue into
of implant puncture or poor visualization of the lower breast, with nonanatomic distri­
the abnormality because of inadequate com­ bution, linear strands, parenchymal bands,
pression . Sonography should be used for all and calcifications. 58 The skin is thickened
masses not definitively imaged mammo­ around the incision sites in the periareolar
graphically and for p al pable masses or thick­ area and inframammary fold. The retroareo­
enings not seen on the mammogram . Ultra­ lar ducts may be interrupted depending upon
sound-directed procedures such as cyst the type of nipple transplantation procedure.
aspirations also can reduce the chance of Fat necro sis is common after breast reduc­
implant puncture. tion , as in other surgical procedures .3 Some­
times presenting as a suspicious palpable
mass, fa t necrosis has many rad iographic
Reduction Mammoplasty appearances (Fig . 25). Often beginning as
vague, patchy areas of mottled architectural
Reduction mammoplasty is a common sur­ distortion, oil cysts may form and calcify,
gical procedure. Indications for breast reduc­ and fibrosis manifested by spiculation may
tion are to bring the breasts into symmetry be seen. "o Calcification is more worrisome
after contralateral breast conservation ther­ when it first appears, but ultimatel y coarse,
apy or mas tectomy and reconstruction; for benign plaques will develop. Oil cysts may
treatment of gigantomastia, with its atten­ form large conglo merate masses that may
dant back pain and strap marks; and to cause clinical concern as palpable findings.
address problems of self-image D DiagnosiS of asymmetric areas of soft-tis­
Although there are many variations, the sue denSity and developing foci of fat necro­
surgical procedure involves a circumareolar sis can be difficult. After reduction mammo­
incision, an inframammary incision, and a plasty has been performed, the sy mmetry of
vertical incision between the two , with re­ the mammographic findings and clinical his­
moval of breast tissue , fat, and skin from a tory can suggest a postsurgical cause.
combination of these vertical and horizontal
incisions (Fig. 24). Breast tissue is removed SUMMARY
predominantly from the lower breast, and
the nipple-areolar complex is brought up­ With widespread use of mammography for
wards. 13 . 58 breast cancer screening, the number of sur­
 EVALUATIOt\ OF THE POSTOPERATIVE BREAST 135

Figure 25. Reduction mammo­

plasty: fat necrosis. A, Craniocau­
dal views show some architectural
disorder after excision of breast
tissue. A soft-tissue density (short
arrow) in the outer aspect of the
left breast most likely represents
fat necrosis, and a grouping of oil
cysts is seen in the outer half of
the right breast (curved arrow).
Linear densities (thin arrows) near
the central posterior breasts sug­
gest scarring related to the infra­
mammary and vertical incisions.
Typical findings in breast reduc­
tion include fat necrosis, paren­
chymal bands, retroareolar ductal
disruption, and alteration of pa­
renchymal architecture. B, More
advanced fat necrosis in another
patient. Thin arcs of calcification
at the rim of oil cysts and heavier
deposition of dystrophic calcifica­
tion are seen bilaterally on these
magnified spot compression cran­
iocaudal views.

gical procedures has also increased. Overlap­ mographic studies will promote the dual goal
ping with radiographic signs of malignancy, of earlv detection of local tumor recurrence
including masses, areas of asymmetric den­ and av"oidance of misinterpreting postopera­
sity and architectural distortion, microcalci­ tive and irradiation changes as malignancy.
fications, and skin thickening, postsurgical Sequential examinations should begin with a
changes may make mammographic evalua­ postoperative preradiation mammogram for
tion difficult. After tumor excision and irra­ residual carcinoma, particularly when micro­
diation where breast alterations are more calcifications have been present, followed by
profound and prolonged, the task of distin­ the baseline postradiation examination at 6
guishing recurrent tumor from scarring or fat months with the next study 6 months later
necrosis is even more challenging. Mammo­ (1 year after initial treatment). Mammograms
grams after breast conservation therapy for of the treated breast may be performed at
carcinoma or after cosmetic surgery require intervals of 6 months until radiographic sta­
correlation with physical findings and the bility has been recognized. Annual studies
surgical procedures that were performed. thereafter are suggested. The contralateral,
Responses of tissue to lumpectomy and unaffected breast should be evaluated mam­
radiation, such as breast edema and skin mographically according to screening guide­
thickening, are most pronounced 6 to 12 lines or clinical concerns.
months after treatment, gradually resolving Mammograms performed after cosmetic
within 1 to 3 years. Carefully tailored mam­ and reconstructive procedures should be cor­
136 MENDELSON

related with the surgical techniques and clin­ 14. Bostwick J III: Reduction mammaplasty. In Plastic
and Reconstructive Breast Surgery, vol 2. St. Louis ,
ical history. Modified views for silicone im­ Quality Medical Publications, 1990, pp 555-1182
plants can maximize visualization of breast 15. Buckley JH , Roebuck EJ: Mamm og raphic changes
parenchyma. Ultrasonography is a useful following radiothe rapy. Br J Radiol 59:337, 1986
complement to mammography in demon­ 16. Cad y B: Choice of operation for breast cancer: Con­
strating the origin of a palpable mass either servative therapy versus radical procedures. 111
Bland K1, Copeland III EM (e ds): The Breast. Phila­
within the implant or the breast parenchyma . delphia , WB Saunders, 1991, pp 753-769
In reduction mammoplasty, distorted archi­ 17. Cardenosa G, Eklund GW: Paraffin block radiog ra­
tecture, parenchymal bands, tissue redistri­ phy following breast biops ies: Use of orthogona l
bution, and fat necrosis should be recog­ views. Radiology 180:873- 874, 1991
18. Cosgrove DO, Bamber JC, Davey J, et al: Color
nized. After mastectomy, myocutaneous Doppler signa ls from breast tum ors. Radiology
reconstruction may be performed. Masses 176:175-180, 1990
that develop within flap reconstructions most 19. Davis SP, Stomper PC, Weidn er N , e t al: Suture
frequently represent fat necrosis, which, calcification mimicking recurrence in the irradia ted
when calcifying oil cysts are seen, may have breast: A potential pitfall in mammographic evalu­
ati on. Radiology 172:247-248, 1989
a characteristic radiographic appearance. 20 . Dempsey PH: Breast sonograph y: His torical p er­
spective, clinical ap plication s and image interpreta­
tion. US Quarterly 6:69-90, 1988
21. Dershaw DO, Chalagssian TA: ~ammogr ap hy after
References
prosthesis placement for augmentati on or recon­
structi ve mammoplasty. Radi ology 170:69-74, 1989
1. Adl er DO: Cltrasound of benign breast co nditions. 22. Dersha,v DO, McCormick B, Cox L, et a l: Differen­
Seminars in Ultrasound, CT an d MR 10:1 06-11 8, tiation of benign and malignant local tumor recurr­
1989 ence after lumpectomy. AJR Am J Roen tgeno l
2. Aspegren K, Holmberg L, Adami HO: Standardi za­ 155:35-38, 1990
tion of the surgical techniqu e in breast-conserving 23. Destouet JM: Mammography of the altered breast.
treatment of mammary cancer. Br J Surg 75:807- 809, Syllabus for the Categorical Course on Breast Im­
1988 aging. ACR, Na shville, TN, 1990, P 78
3. Baber CE, Libshitz H : Bilateral fat necrosis of the 24. Dinner Ml, ARIZ JS: Carcinoma of the breast oc­
breast following reduction mammoplasti es . AJR Am cu rring in routine reduction mammoplasty. Plastic
J Roentgenol 128:508-509, 1977 and Reconstructive Surgery 83:1042-1044,1989
4. Balu-Maestro C, Bruneton J-I'\, Geoffrav A, et al: 25 . Dubois JB , Saumon-Reme M, Gary-Bobo J, et a l:
Ultra son ographic posttreatment foJlow-u~p of breast Tumorectomy and radi ation therapy in early breast
cancer patients. J Ultrasound Med 10:1, 1991 cancer: A report on 392 patients . Rad iol ogy 175:867­
5. Barr LC, Brunt AM, Goodman AG , et al: Cncon­ 87l, 1990
trolled loca l recurrence after treatment of breas t 26 . Eklund GW, Busby RC, Miller SH: Improved imag­
cancer with breas t conservation. Cancer 64: 1203­ ing of the a ugm ented breast. AJR Am J Roen tgenol
1207, 1989 151:4690473, 1988
6. Bassett LW, Gold RH, Mirra JM : Nonneoplastic 27. Endo LP, Edward s N L, Longley S, et al: Silicone
breas t calcifications in lipid cysts: Development after and rheumatic diseases. Semin Arthritis Rheum
exc isi on and pri ma ry irradiation. AJR Am J Roent­ 17:112-118, 1987
genol 138:335-338, 1982 28. Feig SA: The role of ultra soun d in a breas t imaging
7. Beren berg AL, Levene ~B , Tonnesen GL: Mam­ cente r. Seminars in US , CT, and MR, 10:90-105,
mogra phiC evaluation of the postirradiated breast. 1989
In Harris JR, H ellman S (eds): Conservative Man­ 29. Fisher B, Bauer M, Margolese R, et al: Five year
agement of Breast Cancer. Philadelp hia , JB lippin­ results of a transabdo mized clinical trial comparing
cott, 1983, p 265 total mastectomy and segmental mastectomy with
8. Bland KI , Copeland EM Ill: General principles of and without radiation in the treatme nt of breast
mastectomy. In Bland KJ, Copeland EM III (eds): cancer. l\i Engl J Med 312:665-673, 1985
The Breast. Philadelphia , WB Saunders, 1991, p 570 30. Fisher B: Lumpect omy (Segmental Ma stectomy) and
9. Bl and KJ, Cope land EM Ill: H alsted radical mastec­ Axillary Dissection. In Bland KI , Copeland EM III
tom y . In Bland Kl, Copeland EM III (ed s): The (eds): The Breas t. Ph ilad e lphia, WB Saunders Com­
Breast. Phil ade lphia, WB Saunders, 1991, p 594 pany, 1991, p 634
10. Bloomer WD, Berenberg AL, Weissman BN : Mam­ 31. Fleck RS, McI\eese, SiJverbroek :--JA, et al: Conse­
mography of the definitively irradiated breast. Ra­ quences of breast irradiation in pati en ts with preex­
diolo gy 118:425-428, 1976 isting collagen vascular diseases. Int J Rad Oncol
11. Bohm-Velez ~ , Mendelson EB: Computed tom og­ Bio Phys 17:829- 833, 1989
raph y duplex Doppler ultrasound, and magnetic 32. Fowble B, Solin LJ, Schultz OJ, e t al: Breast recurr­
reso nance imaging in evaluating the breast. Semi­ ence following conserva tive surgery and radiation:
nars in US, CT, and MR 10:171-176, 1989 Patterns of failure, prognosis, and pathologic fi nd­
12. Boring CC, Squires TS, Tong T: Cancer s tati stics ings from mastectomy specimens with implications
1991. Atlanta, American Cancer Society, Profes­ for treatment. lnt J Rad Oncol Bioi Phys 19:833- 841,
sional Education Publication, 1991, p 10 1990
13. Bostwick J Ill: Aesth etic breast surgery. In Plastic 33. Gallagher WJ, Cardenosa G, Rubens JR, e t al: Min­
a nd Reconstructive Breast Surgery, vol 1. St. Loui s, imal-volume excision of nonpaJpabIe breast lesions.
Quality Medical Publica tion s, 1990, pp 195-408 AJR Am J Roentgenol 153:957-961 , 1989
 EVA LCATION OF THE POSTOPERATI VE BR EAST 137

34 . Gefter WB , Friedman AK, Goodman RL: The role of 54. Mellink WAM , H endri ks JHCL, Rutgers EJT: Th e
mammograp h y in eva luating patients with early contribution of routine foll ow-up mamm og raph y to
carcinoma of th e breas t for tvlectomv and radiation an ea rl y detec tion of async hrono us con tra la teral
th erapy. Radiology 142:77-86, 1982 ~ breast cancer. Cancer 7:1844- 1848, 1991
35. H all 'FM: Imm ed iate postoperative mammogram for :n Mend elso n EB, Bhag wanani DG, Bohm-Velez M:
fa il ed s urgical exc ision of breast lesions. Radiology Imaging th e breast trea ted by segmen ta l m as tec tomy
180:19- 20, 1991 and irradiati on. Recent Re sult s in Cancer Research.
36. H affty BG, Go ldberg NB, Rose M, et al: Conservative \Jew York Sprin ger-V e rlag, 1990, pp 175-192
surgery w ith rad ia ti on therapy in clinical stage I and 56. Mendelso n EB: Im aging th e post-surgical breast.
II breas t cance r. Arch Surg 124:1266-1270, 1989 Se min a rs in US, CT and MR 10:154-170, 1989
37. Harris JR , Recht A: Conservative surgery and radio­ 57. Mende lso n EB, H arris [Link], Doshi N, Tobon H:
th erapy. 111 Harris JR , Hellina S, Henderson IC, et Infiltratin g lob ular carcinoma: Mammographic pat­
al (eds): Breast Disease, ed 2. Philadelphia, JB Lip­ terns w ith path o log ic correlati o n. AIR Am J Roent­
pincott, 1991 , p 395 gena l 153:265-271 , 1989
38. Harris JR , Schnitt SJ, Connolly JL: Conservati ve 58. Miller C L, Feig SA, Fox JW : Mammographic changes
surgery and radiation therapy for early breast can­ afte r reducti on mammop las ty. AJR Am J Roentgenol
cer. Arch Su rg 122:754·· 755, 1987 149: 35- 38, 1987
39. Harris K:vt , Cos ta-G reco MA, Baratz AB , et al: Th e 59. Mitnick JS , Roses DF, Ha rri s MN, e t al: Calcifications
mam mographic features of the postlumpectomy, o f th e breast after reduction mammoplasty. Surg
pos tirrad iatio n breast. Radi ogra phics 9:253-268, Gyn ecol Obs tet 171409-412, 1990
1989 60. Mitnick J, Roses DF, Harris YIN: Differentiation of
40. H asse ll PR, Oli votto lA, M ueller H A, et al: Ea rl y pos ts urg ica l cha nges fro m ca rci nom a of the breast.
breas t ca nce r: Detection of recurrence a fte r conse r­ Surg Gy necol Obs te t 166:549-550, 1988
vative surgery and radiation therapy. Rad iol ogy 61. Moffa t FL, Ke tcha m AS , Robin so n DS, et al: Seg­
176:731-735, 1990 men ta l mastecto my witho u t rad iotherap y for Tl and
41. Hellman S, Harri s JR : Breast Cancer: Consideration s sma ll T2 breast ca rci no mas. A rch S urg 125:364- 369,
in loca l a nd reg ional treatment. Ra d iology 164:593­ 1990
598, 1986 62 . NIH Conse n su s Conference: Trea tment of early
42. He ster TR, Bost wic k J III : Impl an ts and expa nd ers. stage breast ca nce r. JAMA 265:391-395, 1991
In Bos twick J III: Plastic and Reconstructi ve Breas t 63. Ostee n RT, Conn oll y JL, Recht A, et al: Identificatio n
Surgery. St. Louis, Quality Medical Publis hing , of pa tien ts a t hi gh ri sk for local recurrence after
1990, pp 131-192 conservative surgery and radiation therapy for s tage
43. Heywang SH, Hilbertz T, Beck: Gd-DTPA enhances l or II breast can cer. Arc h Su rg 122:1248-1252, 1987
MR imaging of the breast in patients with p os top ­ 64. Paulus D: Con se rva ti ve treatmen t of breast cancer:
erative scarring in silicon implants. J Com put As sis t Mammography in pa ti en t selection and follow-up .
Tomogr 14:348, 1990 AJR Am J Roen tgenol 143:483-487, 1984
44. Heywang SH, Wolf A, Pruss E, et al: MR imagin g 65. Pau lu s D: Mammogra ph y of the treated breast. In
of the breast with Gd DTP A: Use and limita tions. Feig SA (ed): Breast imaging -American Roentgen
Radiology 171:95, 1989 Ray Socie ty ca tego rical course syllabus. Proceedings
45. Hoeffken W, Lanyi M: The breast followin g surgery of th e 88 th Annual Meeting of the American Roent­
and radiation therapy. 111 Mammography. Philadel­ gen Ray Society, San Francisco, pp 56-57, 1988
phia, WB Saunders, 1977, pp 240-256 66. Peters ME, Fagerholm Ml, Scanlan KA, et al: Mam­
46. Holland R, Hendriks JHCL, Verbee k ALM: Extent, m og rap hic evaluation of the postsurgical and irra­
distribution, and mammographic/ hi sto log ical corre­ diated brea st. Radiographics 8:873-899, 1988
lations of breast ductal carcinoma in s itu . Lancet 67. Rebner M, Pennes DR, Adler DD, et al: Breast
335:519-522, 1990 microcalcifications after lumpectomy and radiation
47. Jackson VP: Sonograph y of malignant breas t disease. therapy. Radiology 170:691-693,1989
Seminars in US, CT, and MR, 10:119-131, 1989 68. Recht A, Connolly JL, Schnitt SJ, et al: Therapy of
48. Kinne OW, Kopans DB: Physical Examinati on an d
in situ cance r. Hematol Oncol CEn North Am 3:691­
Mammography in the Diagn os is of Breast Di sease .
708, 1989
I II Harris Jr, Hellman S, Hend e rson IC, et al (eds):
69. Re cht A, Eberlein TJ, Sadowsky NL: Local recur­
Breast Diseases, ed 2. Philadelphia, JB Li ppincott,
rence following breast conservation. 111 H arris JR,
1991, p 83
49. Kinne DW : Prima rv Trea tm ent o f Breas t Cance r: H ellman S, Henderson Ie, et al (e d s): Brea st Dis­
Surgery. 111 Harri s JR, Hellman S, He nd e rso n Ic' et eases, ed 2. Philadelphia , JB Lippincott, 1991 , pp
al (eds): Breast Diseases, ed 2. Philad e lphi a, JB 54 1--546
Lippincott, 1991, pp 347- 373 70. Recht A, Ha yes DF : Loca l recurrence fo ll ow ing ma s­
50. Leibman AJ, Kru se B: Breas t Cancer: Mam mog rap hic tectom y . In Harris JR , H e llm a n S, H e nderson IC, et
and so nog raphi c finding s after augm e nta ti o n ma m­ al (ed s): Breas t Diseases, ed 2. Philade lphia, JB
moplasty. Radi ology 174:195-1 98, 1990 Lippincott, 199 1, p 529
51. Libshitz HI, Montague ED, Pau lu s DD: Calcifi ca ti o ns 71. Robertson JM, Clarke DH, Pa vzner MM , e t a l: Breas t
and the therapeuticall y irrad ia ted breast. AJR A m J co nserva tion therapy-seve re breas t fibro s is a fter
Roentgenol 128: 1021-1 025, 1977 radiation therap y in patients wi th collagen vascular
52. Libshitz HI, Mon ta gue ED, Pau lu s DD : Ski n thick­ di sease. Cancer 68:502-508, 1991
ness in the therapeutically irradiated breast. AJR Am 72. Rosenth a l E: Man y w ith breas t cancer have limited
J Roentgen ol 130:345-347, 1978 surgery. New Yor k Tim es Jul y 21, 1991, pp AI , A 18
53. McCraw JB, Cramer AR, Horton C E: Breast Recon ­ 73. Sadow sky NL, Semine A, Harris JR: Breast imag in g:
struction Foll ow in g Mastec tomy. 111 Bland KI, Cope­ A critical aspect of breas t con ser ving treatment.
land EM III (eds): Th e Breas t. Philadel phia, WB Cancer 65:2113-2118, 1990
Saunders, 1991, pp 656-693 74. Sadowsky N L, Semine A: Good m a mm og raphy
138 MENDrLSON

finds postop cancer recurrence. Diagnostic Imaging mammogram for failed surgical excision of breas t
October:100-106, 1990 lesions. Radiology 178: 159-162, 1991
75. Schapira DV, Urban \I: A minimalist policy for 85. Stigers KB, King JG , Davey DO, et al: Abnormalities
breast cancer surveillance. JAMA 265:380-382, 1991 of the breast caused by biopsy: Spectrum of mam­
76. Sickles EA, Herzog KA: Mammography of the post­ mographic findings. AJR Am J Roentgenol 156:287­
surgical breast. A)R Am J Ro en tgenol 136:585-588, 291, 1991
1981 86. Stomper Pc' Gelman RS: .'vlammography in symp­
77. Sickles EA: Periodic mammographic foll owup of tomatic and asymptomatic patients. Hemat ol Oneol
probably benign lesions: Results in 3,184 consecutive Clin North Am 3: 611-640, 1989
cases. Radiology 179:463-468, 1991 87. Stomper pc, Recht A, Berenberg AL, et al: Mam­
78. Silverstein MJ, Gierson ED, Gamagami P, et al: mographic det ection of recurrent cancer in the irra­
Breast cancer diagnosis and prognosis in women diated breast. AJR Am J Roentge nol 148:39-43, 1987
augmented with silicone-gel fi lled implants. Cancer 88. Sunderland MC, :-vlcGuire WL: Prognostic indicators
66:97-101, 1990 in in vasive breast cancer. Surg Clin i\o rth Am
79. Singletary SE, McNeese M: Segmental mastectomy 70:989 1003, 1990
and irradiation in the treatment of brea st cancer. 89 . Tabar L, Dean PB: Teaching Atlas of Mammography.
Am J Clin Oncol 11:679-683, 1988 Stuttgart, Georg Thieme Verlag, 1983, pp 88-90.
80. Solin LJ, Fowble BL, Schultz OJ, et al: Definitive 134-5
irradiation for intraductal carcinoma of the breast. 90. Veronesi U, Banfi A, Sa lva dori B, e t al: Brea st
lnt J Rad Oneol Bio Phys 19:843-850, 1990 conservation is the treatment of choice in small
81. Solin L), Fow ble BL, Schultz OJ, et al: The detection breast cancers: Long-term results of a randomized
of local recurrence after definitive irradiation for trial. Eur J Cancer 16:668-670, 1990
early stage carcinoma o f the breast: An ana lysis of 91. Veronesi U: Quadrantectomy. In Bland KI, Copeland
EM III (eds): The Breast. Philadelphia, WB Saunders,
the results of breast biopsies performed in previ­
1991 , P 631
ously irradiated breasts. Cancer 2497-2502, 1990
92. Wickerham DL, Fisher B: Surgical treatment of pri­
82. Souba WW, Bland KI: Indications and techniq ues mary breast cancer. Semin Surg Oncol 4:226-233,
for biopsy. In Bland Kl, Copeland fM III (eds): The 1988
Breast. Philadelphia, WB Sa unders, 1991, pp 534 93. Wolfe J: Xeroradiography of the Breast, ed 2. Spring­
550 field , Charles C Th omas, 1983, p 20 .
83. Stehlin is, delpolyi PO, Greeff Pi, et al: A ten year 94. Wolfe J: Xe roradi ograp hy of the Breast, ed 2. Spring­
study of partial mastectomy for ca rcin oma of the field , Charles C Thomas, 1983, p 405
breast. Surg Gynecol Obstet 165:191-198, 1987 95. Wolfe J: Xeroradiograp hy of the Breast, ed 2. Spring­
84. Stein MA, Karlan MS: Immedia te postoperative field, Charles C Thom as, 1983, p 419

Address reprinl requests 10

Ellen B. Mendelson, MD
Department of Radiology
The Western Pennsylvania Hospital
4800 Friendship Avenue
Pittsburg h, PA 15224