Breast Dowlatshahi,
Kambiz
MD
#{149} M.
Nonpalpable of Stereotaxic Fine-Needle
Lisa
terms:
Biopsies, #{149} Breast
biopsy, 00.125 00.31, 00.32
Radiology
I
From
1991;
(L.F.K.),
technology neoplasms,
and
Breast, diagnosis, #{149}
181:745-750
the Departments
(K.D.), Pathology
of General
(M.L.Y.),
Radiology
Surgery
Cytopathobogy
(P.M.J.),
Rush-Presbyte-
nan-St Luke’s Medical Center, 1653 W Congress Pkwy, Jelke Bldg. Rm 769, Chicago, IL 60612. From the 1989 RSNA scientific assembly. Received revision
dress
March 8, 1991; received July
reprint
requests
2 Current address: Biology, University
0 RSNA,
revision requested 8; accepted July
D
May 7; 23. Ad-
of Molecular Chicago.
F. Kluskens,
and
Reynolds
ESPITE
MATERIALS From
June
mographically for
1991
See also the editorial byjackson (pp 633-634) in this issue.
#{149} Larry
advances in mammographic technique, it remains difficult to correctly predict the benign on malignant nature of mammographically detected breast lesions. Even with these improvements, most lesions detected nadiographically and recommended for biopsy are benign at subsequent histologic examination. The positive predictive value of mammography in recent reports ranges from 14% to 38% (1-10). In 1986, steneotaxic fine-needle-aspiration (FNA) cytologic study of occult breast lesions was introduced into the United States in an attempt to overcome this pnoblem (11). Results from this and other studies showed that the procedure has a false-negative rate of 5%-14% and an inadequate sampling rate of 10%-26% (12-15). Also, most pathobogists in the United States have bess experience in breast cytology than in histology, and cytologic diagnoses are received by clinicians with less confidence. In 1988, we added fine-needle conetissue biopsies to FNA cytologic assessment in our diagnostic evaluation of nonpalpable, mammographically detected breast lesions to determine whether an acceptable and safe sobution could be found to reduce the rate of unnecessary open-breast biopsies. A mechanized fine-needle cone biopsy device (Biopty; Bard Unobogicab, Coyington, Ga) that has been used under sonographic control by physicians to procure tissue samples of suspicious liven, chest, and prostate nodules (1619) was used with our stereotaxic approach. This report presents the results of our experience in 250 such cases.
to K.D. Department of Chicago,
MD2
MD, PhD
#{149} Peter
M. Jokich,
MD
Breast Lesions: Findings Needle-Core Biopsy and Aspiration Cytology’
Two hundred fifty mammographically detected nonpalpable breast lesions suspicious for malignancy in women who underwent routine screening mammography were stereotaxically localized. Fine-needleaspiration (FNA) cytologic specimens and needle-core biopsy specimens were obtained before open biopsy in every case. Seventy-six lesions (30.4%) were malignanL Sixty-three (83%) of these 76 cancers were 1 cm long or smaller. Needle-core biopsy alone was used to diagnose conclusively 41% (ii = 31) of these cancers, while FNA cytologic study alone was used to diagnose 32% (n = 24). No false-positive results occurred with either test. The same diagnosis was reached in 54% (n = 41) when the combined results of both needle tests were considered. In applying the two needle tests to 125 mammographically defined low-suspicion lesions, 85 (68%) were found to be benign by means of either one or both needle tests; there was one lobular carcinoma in situ. By applying this algorithm, 85 (34%) of 250 patients with abnormal mammograms, or one-third of all patients recommended for open biopsy, might have avoided surgery. Index
Yaremko,
Imaging
carcinoma
AND
METHODS
1988 to May 1990, detected lesions in 242
women
250 mamsuspicious
study
was
granted by the Institutional Reand informed consent was obtained from each patient. For the punposes of this study, every abnormal mammogram depicting either soft-tissue masses or microcalcifications was reviewed by two investigators (K.D., P.M.J.) and was classified as having a high, intermediate, or low degree of suspicion for malignancy as follows: (a) A high index was indicated by a solid soft-tissue mass having highly irregular or stebbate borders; fine-needle-like and linear branching microcalcifications, either clustered or scattered; extremely polymorphic clustered microcalcifications; or an irregularly marginated solid mass containing suspicious or indeterminate microcalcifications. (b) An intermediate index was indicated by a solid soft-tissue mass with partially irregular or obscured borders or indeterminate microcalcifications. (c) A bow index
view Board,
was
indicated
by
a 0.5-1.5-cm
well-defined
mass, a 1-2-cm asymmetric clustered calcifications that
opacity, and were primarily rounded and rather uniform in size. If comparison mammograms were available, lesions were moved up or down on the scale of suspicion, depending on whether they were new or stable for 2-3 years; the latter group would be followed by means of mammography. The size limit of lesions for needle biopsy was lowered from 1 cm to 0.5 cm with experience. During the batter
part
of the
of patients with for malignancy
period,
approximately
mammograms
20%
suspicious
underwent
additional
di-
agnostic mammography consisting of spot-compression magnification views and/or ultrasound. The details of stereotaxic localization and FNA cytologic studies have been previousby
described
(12).
To be
lesion was stereotaxically the mammographic unit Fischer
Imaging,
Denver).
brief,
each
revisualized (Mammotest; The
skin
on
over
the puncture site was anesthetized with 1 % lidocaine, and a double-lumen coaxial needle (Cook Urobogical, Spencer, md) was inserted into the breast. The accuracy of needle localization was verified with stereoradiographs obtained before samplings. At least two and sometimes three
undergoing
routine screening mammography prospectively studied. Permission
were for this
Abbreviation:
FNA
=
fine-needle
aspiration.
passes for cytologic assessment were made with the inner 20-gauge Franzen-type needle. The core biopsy procedure performed with a 20-gauge Biopty needle fobbowed immediately after the cytologic aspiration. Again, two and sometimes three samples were obtained through the same puncture sites. Stereoradiographs of the needle in the lesion were obtained to document the part of the lesion obtained for biopsy.
In September
1989,
after
130
(120
cases).
gun had a lighter needle penetration, 17 mm.
The
The
spring, and
result
was
1
Histologic
to Radiographic
Correlation
Mammographically
Suspicious
Appearance
Nonpalpabbe
Suspicion
Index
of
and Type of Lesion
High or intermediate Mass
(n
=
No.
49(20) 76 (30)
10(6) 50 (29)
39(51) 26 (34)
80 34
93
84
9 (12)
9
(37)
32 (13) 250(100) in parentheses
a substantial
confirm the location of the wire. Only in 10% of cases were orthogonal views obtamed
for
following
(a) Two
intravenous
were
FNA
cytologic
biopsy by two
studies;
M.L.Y.,
without
knowledge
biopsy
Open-biopsy
to the Surgical where they were and processed for study.
fresh
specimens
needle-core
and
nee-
biopsies)
of the results
in a weekly
to standard
al-
specimens were reviewed of us (L.F.K., FNA cytologic
review
ries, and radiographs tobogic specimens ing
95%
of open
of cases,
histo-
by the authors. Cywere evaluated accord-
cytologic
criteria
(20,21)
and classed as negative (benign epitheliab elements only), atypical, suspicious for carcinoma, positive for carcinoma, or insufficient for diagnosis. The criterion for adequacy of cytologic specimens was more than six epithelial clusters of more than 10 cells each. Needle-core biopsy specimens were evaluated for size (representative of the size of a needle core, or approximately 7 mm bong) and presence of breast epithebial elements and stromal tissue.
The
epitheliab
elements
were
as-
sessed as benign, hyperplastic with or without atypia, suspicious, or malignant according to standard histopathobogic criteria (22). A needle-core sample was classed as insufficient if the histologic specimen was less than 7 mm long (one piece or total length of several pieces) or extensively fragmented, or if crush artifact precluded visualization of intact epitheliab cells. Finally, the findings of open biopsy were reviewed and results of the three modalities
compared.
After cytologic and needle-core biopsy specimens of the suspicious lesion were obtained, the location of the lesion in the breast was doubly marked with 0.3 mL of methylene blue dye injected through the localizing needle. A hook wire was inserted through the same needle. In 90% of cases, stereoradiographs were obtained to
746
#{149} Radiology
was
lesions
reasons:
Department.
with
separate
the
specimens
All
(%)
(48)
30 (17) 174(100)
2 (3) 76(100)
6 30
im-
Pathology
dle-core blindly
Value
are percentages.
cohob, and Papanicolaou staining. The needle-core biopsy specimens were placed in 50% Bouin fixative solution and submitted for routine histopathobogic processing and evaluation by members of the Surgical
Pathology Department examined, sectioned, routine histopathobogic
Predictive
faster speed of sampling notch of
fixation
sent
Positive
Biopsy Malignant
been localized in the same breast, and (b) the accuracy of localization was in doubt (eg, small lesions in a dense breast or a far posterior location). Patients were then taken to an operating room for excision of the lesion with the use of local anesthesia with or without
processing,
of Open
of
125)
Microcabcification Overall Note-Numbers
for 250
Benign
Mass
Monopty
of Suspicion
Lesions
Lesions
Microcalcification Low(n=125)
gun towas this
provement in the specimen quality and adequacy rate. Cytologic specimens were sent immediately to the cytopathology laboratory for cytospin
and Index
Breast Results
Mammographic
proce-
dures were performed, a new biopsy (Monopty; Bard Urobogical) that was tally disposable became available and exclusively used for the remainder of investigation
Table
had
sedation.
infrequently
General
anesthesia
given.
RESULTS Two
fifty lesions in 242 examined. The average age of the patients in this study was 53 (range, 30-85) years. Seventy-six malignant (30.4%) and 174 benign (69.6%) lesions were verified with open biopsy, for a positive predictive value of 30.4% for mammography in this series. Fifty-four cancers (71%) were invasive, and 22 (29%) were in situ. Overall, 108 lesions (43.2%) were suspicious microcalcifications, of which 29 (26.9%) were histologically proved to be malignant. One hundred forty-two lesions (56.8%) were soft-tissue abnormalities, of which 47 (33%) were malignant. The distnibution of lesions in the three mammographic categories and subsequent histologic findings are shown in Table
women
1. For
hundred
were
practical
reasons and simplicity mammognaphic lesions of high and intermediate suspicion were grouped together. This arbitrary division shows that it is possible to separate abnormal mammograms into two groups with a highly significant (x2 test; P < .001) difference in positive predictive value for malignancy. Half of all lesions (ii = 125), however, were in the low-risk mammographic category, with a prac-
of presentation,
ticab management outcome for needle tests. Table 2 shows the results of FNA cytologic assessment, needle-core biopsy, and both tests against the re-
sults of open biopsy. As expected, greater number of definitive diagnoses, either benign or malignant,
a
were
reached with needle-core biopsy than with FNA cytologic assessment. Twenty-four (32%) of 76 malignant lesions were conclusively diagnosed with FNA cytologic study, and 31 (41 %) were conclusively diagnosed with core biopsy. There was no falsepositive result. The sensitivity and specificity of these two tests were determined by including atypical, suspicious, and malignant lesions as positive and benign lesions as negative. Insufficient biopsy material for histologic diagnosis was regarded as no test. Thus, the sensitivity was 86%, specificity was 72%, and rate of insufficient results for FNA cytologic assessment was 24%. The same results for
core biopsy were 71%, 96%, and respectively. When the Biopty needle was replaced with the technically improved Monopty device after the first 130 cases, the rate of insufficient tissue sampling was reduced from24% to 8%. Table 2 also shows the combined results of FNA cytologic study and needle-core biopsy in 250 lesions. The 17%,
following nancy,
were
and chosen:
criteria for indeterminate Benign
benignity, maligdiagnoses lesions
were
those for which either both needle tests were negative or one was negative and the other was deemed insufficient. Atypical lesions were those for which either cytologic study or core biopsy showed dysplastic cells or atypical hyperplasia. Suspicious besions were those for which either one or both tests showed cells or tissue suggestive of but not definitive for malignancy. Malignant lesions were those for which either one or both needle lignancy.
tests revealed Employing
(54%) of 76 malignant conclusively diagnosed
evidence of masuch criteria, 41 lesions with
were a sensi-
December
1991
Table 2 Correlation
of Results
of FNA Cytologic Assessme Suspicious Breast Lesions
250 Mammograph.ically
nt,
Needle
-Core Biops y, and
Both
Techniques Open
Needle
Findings
(n
Cyto = 250)
Benign Atypical
99 41
Suspicious
25
Malignant Insufficient Note-Both
(n
both cytologic
155 11 10 31 43
assessment
and
needle-core
Table 3 Mammographic and Pathologic Features Cytologic Assessment and Needle-Core Lesion No.
Mammographic Appearance
-
Both = 250)
Cyto (n
126 39 27 41 17 biopsy,
of Malignant Biopsy
(n
=
biopsy
Missed
I
IDC
-
-
-
DCIS DCIS
-
-
IDC
I
-
Microcalcifications
I
-
-
-
8
Microcalcifications Microcalcifications
-
-
DCIS DCIS DCIS DCIS
9
Microcalcifications
I
I
DCIS
irregular
9-mm 5-mm
irregular mass stellate mass
4 5
Microcalcifications Microcalcifications
6
7
=
mass
findings, I = insufficient carcinoma in situ, IDC
Cyto
Correlation
of Combined Classification
FNA Cytologic
Assessment
in 125 Abnormal
Mammograms
and Needle-Core
Biopsy
Findings
Benign
with Lesions
of Low-Suspicion (n
FNA Cytologic and Core Biopsy (n = 125)
Benign (n = 115)
=
Classification 125) Malignant
(n
10)
=
85 20
85
0
18
2
Suspicious
8
Malignant Insufficient
6 6
6 0 6
2 6 0
Atypical
for diagnosis
tivity of 89%, specificity of 75%, and insufficiency rate of 7%. Sixty-three (83%) of 76 cancers were 1 cm or smaller in diameter at mammography. None of the 13 cancens banger than 1 cm was found to be benign by means of either needle test. Compared with FNA cytologic examination, needle-cone biopsy abbowed a superior (26 vs 15) diagnosis of malignancy when a mass was seen on the mammogram. On the other hand, FNA cytologic testing scored better (nine vs six) in the diagnosis malignant microcabciflcations. Also, (45%) of 29 malignant micnocabciflca-
of 13
cytologic
(n
9 15 16 24 12 assessment
were ductab nine lesions ameter.
=
Both
76)
(n
19 7 8 31 11
=
76)
8 10 16 41 1
only.
carcinoma were 1 cm
in situ. on less
All in di-
Application opsy method
Histologic Combined Assessment
=
Core
Cyto = 76)
(n
DISCUSSION
tissue for diagnosis. = invasive ductal carcinoma.
Table 4 Histologic
Both = 174)
Table 4 shows the results of the combined needle tests applied to lowrisk lesions. The insufficiency rate was 4.8% (six of 125) with no false-negative results but one case of lobular carcinoma in situ. The sensitivity was 100%, specificity was 78%, and negative predictive value was 100%.
Histologic Assessmentt
I
6-mm
(n
118 29 11 0 16
Both
with
Core Biopsy Findings*
only,
-
1
174)
136 4 2 0 32
needle-core
Lesions
=
in
Malignant
Core
174)
=
90 26 9 0 49
Core
Cytologic Assessment*
2 3
negative t DCIS = ductal *
(n
B iopsy
Biopsy
Benign
Core = 250)
24 61
for diagnosis =
Biopsy
Resul ts of Open
with
tions were falsely reported as benign with needle-core biopsy; only three (10%) of 29 were falsely reported as benign with FNA cytologic assessment. Histologically proved fibnoadenomas presented a sampling problem: These lesions shifted by about 5 mm as judged on postinsention radiographs. Table 3 presents the details of nine lesions missed with both cytologic examination and needle-cone biopsy. Six of the nine were micnocalciflcations. Only one of six lesions first seen as microcalcifications was invasive ductal carcinoma; the remaining five
of the needle-core biin conjunction with the steneotaxic device (Mammotest) brings a new dimension to the diagnosis of nonpalpabbe breast lesions suspicious for malignancy. The current average ratio of benign to malignant lesions for open biopsy of suspicious mammographic lesions in the United States is 3:1 (1,4,9). This is partly due to the unfavorable medicolegal risk exposure in this country beading to increased frequency of biopsy of bow-suspicion lesions. Neverthebess, the likelihood of abnormal mammograms being predictive of malignancy ranges from 10% in asymmetric panenchymal opacities and moderately suspicious microcalcifications to 50%-70% for stelbate masses and branching microcalcifications (5). The lesions of the first category in the “lighter shade of gray” may be safely followed with interval mammography, as recently reported in a webbconducted study (10). The number of unnecessary excisional breast biopsies may also be reduced by the application of stereotaxic fine-needle biopsy in conjunction with mammography. Azavedo et al reported only one cancen missed in 2,005 lesions examined in this way (23). Other investigators, also from Europe, reported favorably on their experience with steneotaxic fine-needle biopsy (14,24). Our results Ra1inlnov
#{149} 747
b.
a. Figure 1. magnification,
same
size
(a) Microscopic
appearance of a 1.5-cm fibroadenoma x 40.) (b) Microscopic appearance of a 1 .4-cm as that in a. (Original magnification, x40.)
show that stereotaxicably guided needle-cone biopsy, either alone on in conjunction with FNA cytologic assessment, is not a substitute for open biopsy in every instance. If selectively applied to mammographicalby defined low-risk lesions in which the prior probability of malignancy is less than 10% (Table 1), however, an excelbent degree of diagnostic reliability (100% sensitivity and negative predictive value) may be achieved (Table 4). We found the most serious limitations in the method to be (a) target size and mobility and (b) technique.
obtained by means invasive ductal carcinoma
ABNORMAL
Size
and
Mobility
(25).
All nine
ble
false-negative
cases
748
#{149} Radiology
needle. (Original a needle the
(n
INTERMEDIATE-
& 111CR-SUSPICION = 125)
LESIONS
UNif COST
t
,i
RE
BIOPSY
$11100
.,..5,
. NEEDLE
.
LOCALI
m
± NEEDLE
. . -.
:
5’..
$100 ‘
MAMMOGRAPHY (6.12 MONTHS)
BIOPSY
(LUMPFrrOMt
$2,100
I $102,000
$152,100
I
GRAND TOTAL
j
AvERAGE
$254,000 NEEDLE
I OPEN
BIOPSY
COST
P
CASE
Low-suspicion IntermediateFigure lesions.
EXCISI( 15
+ .#{149} OPEN
!
(Ta-
3) occurred among lesions first seen as either microcalcifications on a soft-tissue mass smaller than 1 cm, suggesting that mammographicably highor intermediate-suspicion besions of this size should undergo open biopsy. Also, seven of nine missed lesions were in situ carcinomas that might have shown a change in appearance at interval mammography if they had not been excised. Needle-cone biopsy can definitely be used to diagnose invasive carcinoma (Fig ib), but it is unlikely to allow confirmation of ductab carcinoma in situ. Suspicious microcabciflcations, which constituted 43.2% (n = 108) of all lesions in our series and 38% (n = 29) of malignant lesions, presented a special problem. This subgroup of breast lesions represents the earliest form of detectable breast cancer. Malignant micnocabcifications represent islands of cancer (ducts on lobules) less than 1 mm in diameter and therefore equal to the tip of cytology
urr COST
,!‘ .
r1i:
TOTAL
Seventy-eight percent of all lesions and 83% of cancers were equal to on less than 1 cm in diameter, compared with 40% reported a decade earlier
.. ,
Monopty biopsy with
MAMMOGRAMS
.
LSSIONS 123)
TOTAL SUB.
Target
of core biopsy with a 20-gauge obtained by means of core
2. Proposed Related costs
mab findings,
insuff
algorithm for each =
insufficient
mammogram & high-suspicion
$2,000
$3,000
mammogram
for management of nonpalpabbe diagnostic alternative are in the
mammographically shaded areas. Abnorm
depicted
abnor-
=
findings.
and cone biopsy needles. It is of interest that FNA cytologic assessment scored better (nine malignant and six suspicious lesions) than the needlecore biopsy (six malignant lesions and one suspicious lesion) in 29 malignant microcalcifications. We speculate that the reason for this difference is the to-and-fro motion effect of the Franzen needle that breaks down the supportive structures and frees the cells that are aspirated into the needle. We believe that needle-core biopsy and FNA cytologic assessment complement each other in the diagnosis of suspicious microcalcifications. Our experience shows that the correct diagnosis of these minute lesions is more likely with multiple passes of the needle (both types) into the besion, which we practice and strongly
recommend.
If the
suspicion for spite adequate
cancer remains high negative cytologic
cone biopsy biopsy
samples,
should
Adequate
of the
antee
fixation
about
to be
open immonot
lesion
guan-
when
by
the
it is needle.
in the location of the lesion suggestive of a fibroade-
as opposed is relatively
during
does
penetrated
which thebess,
an and
breast
of the
noma,
negative target
then
deand
be performed. compression
bilization
The shift is highly
mammographic
the
lesion
to a carcinoma, immobile. Nevershift
tissue sampling. may occur more a core
biopsy
results
in false-
Missing frequently sampling
the be-
cause the needle is mechanically fined into the lesion and the “feel” of tissue resistance experienced by means of December
1991
manual
penetration
logic
during
assessment
Technical
FNA
cyto-
is lost.
Aspect
There was a substantial improvement in the specimen quality noted by our
pathologist
when
the
fragmentation
as webb
surgical
Biopty
needle was replaced with the Monopty device midway through study. The prevalence of specimen as the
the
prohibiting
multiple
micro-
sampling
of
the target. Also, a small incision is necessary for skin penetration. The final choice of the needle may depend on the lesion size: the bigger needles for 1-2-cm soft-tissue lesions and the finer needles for microcalcifications. Diagnostic Mammograms
Plan
for Abnormal
Mindful of the current increasing liability exposure in the United States, we propose the following algorithm. We strive to reduce the chance of missing
breast same
the
diagnosis
cancer as the
of an
to 1%-2%, reported
early
which
is the
prevalence
for
open biopsy (6,27-29). The threshold of this risk acceptance may be raised with experience on in individual patients who refuse to undergo open biopsy. The algorithm is summarized in a flow chart Mammographically
mediate-risk ommended ization and FNA
cytologic
(Fig
2). high-
on inter-
lesions are generally necto undergo needle localopen biopsy. Optional assessment
performed
at the time of localization with immediate reporting (within 1 hour) serves two purposes: (a) A positive diagnosis of malignancy will assist the surgeon in performing a complete bumpectomy with tumor-free margins, thus avoiding the risk of a difficult and unsatisfactory ne-excisionab operation. This is particularly important in the present-day practice of breast consenvation. (b) In cases of two mammoVc
me 181
#{149} Number
3
management.
An
asymptom-
atic woman whose screening mammogram reveals a low-risk lesion is advised to undergo stereotaxic localization,
scopic crush effect decreased markedly, beading to a sample adequacy rate of 92% for Monopty compared with 76% for Biopty. The cone biopsy specimens procured with a 20-gauge needle (Fig 1) were generally adequate for histologic interpretation. Parker et ab (26) reported that a more satisfactory specimen sample was obtamed with 14-gauge Biopty needles. After completion of this study, we began using 18-gauge needles, as nonprotocol pathologists in our institution prefer to work with larger biopsy specimens. We believe that larger needles cause more bleeding, thus
graphically separate, yet suspicious, lesions in the same breast on extensive suspicious microcalcifications that exdude breast conservation, positive needle diagnosis of lesions will help the surgeon and the patient decide on
FNA
cytologic
assessment,
to undergo
mography within such an algorithm to the 125 bow-risk 85 lesions
would
interval
have
been
Shott,
versus
2.
is given
in Figure
2. Although
the
biopsy
for
250
cases
6.
advice;
Roshni
Patel,
AL, Schwartz AS. Clinically
GF, Seig SA, Paoccult breast be-
sions: localization and significance. Radiology 1987; 162:167-170. Tinnemans JGM, Wobbes TH, Holland R, et
Rowen
JP, Bassett
55, Simon
IS, Brown
DS.
biopsy for mammoin 561 patients. Can
Silverstein MJ, Gamagami P, Rosser RJ, et al. Hooked-wire directed breast biopsy and overpenetrated mammography. Cancer 1987; 59:715-722. Ciatto 5, Cataliotti
lesions
Z, Distante
detected
V.
Nonpal-
with mammogra-
phy: review of 512 consecutive ology 1987; 165:99-102. Hall FM, Storella JM, Sibverstone
cases.
Radi-
DZ,
Wy-
shak G. Nonpalpable breast lesions: recommendations for biopsy based on suspi-
7.
cion of carcinoma at mammography. Radiology 1988; 167:353-358. Hermann G, Janus C, Schwartz IS, Krivisky
B, Bier 5, RabinowitzjG. breast lesions: accuracy
8.
diagnosis.
Nonpalpable of pre-biopsy Radiology
mam-
mographic 323-326. Papatestas Tsevdos
AE, Herman C, Lesnick G.
D, Hermann G, Surgery for non-
palpable
breast
Arch Surg 1990;
lesions.
1987;
165:
125:399-402.
9.
of ab-
Acknowledgments: We gratefully appreciate the collaboration of Steven D. Bines, MD, Daniel J. Deziel, MD, Alexander Doolas, MD, Steven G.
to
Susan
BA, and Debra Babich, MD, assistance; and Jan Nunnalby help.
Needle-guided breast graphic abnormalities Surg 1986; 29:287-288.
cost
normal mammograms with that for the same number if handled according to the proposed algorithm. This cost savings is attained with no risk to patients. Also, one-third of women derive superior cosmetic results, and their unscarred breasts could be evaluated more satisfactorily at subsequent examinations. On the basis of the result of this and other studies, we believe that steneotaxic needle biopsy of nonpalpable breast lesions will play a significant robe in the future diagnosis and management of breast cancer. It will reduce the number of unnecessary breast biopsies and the associated cost and will encourage more women to participate in screening mammographic surveillance for early detection of breast cancer. U
for statistical
Rosenberg tachefsky
pable
of individual needle biopsy may be one-third that of excisional biopsy ($1,100 vs $3,000), the average cost for the low-risk lesions with associated open biopsies in 40 of 125 cases would raise the average diagnostic cost of this group to $2,000 versus $3,000 for high- and intermediate-risk lesions. Nevertheless, a 15% cost reduction is achieved if we compare the cost of excisional
PhD,
patients
165:523-529.
3.
4.
cost for stereoversus that of outat our institution groups of lesions
their
We also thank
al Mammographic and histopathologic correlation of nonpalpable lesions of the breast and the reliability of the frozen section diagnosis. Surg Gynecol Obstet 1987;
Benefit
The approximate taxic needle biopsy patient open biopsy in 1991 for the two
encouraged
in this study.
References
5.
Cost
who
BA,Julee Bangert, for their technical for her secretarial
diag-
of either
MD,
participate
mam-
6-12 months. If had been applied cases in this series,
nosed as benign by means one on both needle tests.
Wool,
1.
and cone biopsy. If both samples are reported inadequate or one is abnonmab (atypia, suspicious, on frankly malignant), open biopsy should be penformed. If one or both needle samples are adequate and benign, the patient is advised
Economou, MD, Keith W. Millikan, MD, Theodore J. Saclarides, MD, Edgar D. Staren, MD, PhD, Thomas R. Witt, MD, and Norman L.
10.
Franceschi
D, Crowe
J, Zollinger
R, et al.
Biopsy of the breast for mammographically detected lesions. Surg Gynecol Obstet 1990; 171:449-455. Sickles EA. Periodic mammographic
bow-up
of probably in 3,184 consecutive
benign cases.
fol-
lesions: results Radiology 1991;
179:463-468.
11.
Dowlatshahi
12.
Bibbo M, Dawson PJ. Cytologic diagnosis of occult breast lesions using stereotaxic needle aspiration. Arch Surg 1987; 122: 1343-1346. Dowbatshahi K, Gent HJ, Schmidt R, Jokich
K, Jokich
PM, Schmidt
PM, Bibbo M, Sprenger
E.
R,
Nonpalpable
breast tumors: diagnosis with stereotaxic localization and fine-needle aspiration.
13.
diobogy 1989; 170:427-433. Svane G, Silfversw#{228}rdC. dle biopsy of nonpalpable
Stereotaxic neebreast lesions:
cytologic and histologic findings. Acta diol [Diagnj(Stockh) 1983; 24:283-288.
14.
Lofgren
M, Andersson
Stereotactic
logic diagnosis sions. AJR
15.
Jackson
I, Lindholm
fine-needle
Ra-
aspiration
of nonpalpable
Ra-
K. for cyto-
breast
le-
1990; 154:1191-1195.
VP, Bassett
LW.
Stereotactic
fine-
needle aspiration biopsy for nonpalpable breast lesions. AIR 1990; 154:1196-1197. 16.
Anderssen
T, Erikson
B, Lindgren
PG, lvi-
lander E, Obeng K. Percutaneous ultrasonography-guided cutting biopsy from liver metastases of endocrine gastrointestinal tumors.
Ann
Surg
1987;
206:728-732.
Radiobosrv
#{149} 749
17.
18.
Goralnik CH, O’Connell DM, EL Yousef SJ, Haaga JR. CT-guided cutting needle biopsies of selected chest lesions. AJR 1988;
22.
151:903-907.
23.
Ragde
H, Aldape
trasound-guided gun superior
Urology
1988;
19.
32:503-506. Kaye KW, Horwitz
20.
sound-guided prostate biopsies using new automatic gun: analysis of 100 consecutive cases. J Endourol 1989; 3:155-162. Kline TS. Masquerades of malignancy: a
21.
review of 4,241 aspirates from the breast. Acta Cytol (Baltimore) 1981; 25:263-266. LinskJA, Franzen S. Clinical aspiration cytology. 1989.
2nd
ed.
24.
CA.
Transrectal
Philadelphia:
ultra-
Lippincott,
York:
histopa-
27.
Churchill-
G, Auer in 2594
G. Stereotactic mammographi-
cally detected non-palpable lesions. Lancet 1989; 1:1033-1036. Gent HJ, Sprenger E, Dowbatshahi K. Steneedle
cal diagnosis
26.
Diagnostic New
1987.
Azavedo E, Svane fine needle biopsy
reotactic
25.
J.
of the breast.
Livingstone,
HC, Bagley CMJr. UIprostate biopsy: Biopty
to aspiration.
Page DL, Anderson thobogy
localization
of occult
breast
and
28.
cytobogi-
lesions.
Ann
Surg 1986; 204:580-584. Moskowitz M. Significance of minimal breast cancers detected on screening. AIR 1981; 136:735-738. Parker SH, Lovin JD, Jobe WE, Burke BJ,
Hopper KD, Yakes WF. Nonpalpabbe breast lesions: stereotactic automated large-core biopsies. Radiology 1991;
29.
Landercasper J, Gundersen SB Jr, Gunderson AL, Cogbill TH, Travebli R, Strutt P. Needle localization and biopsy of nonpal-
pable lesions of the breast. Surg Gynecol Obstet 1987; 164:399-403. TinnemansjGM, Wobbes T, Hendriks JHCL, Van der Sluis RF, Luberes EJC, deBoer l-IHM. Localization and excision of nonpabpable breast lesions: a surgical evaluation of three methods. Arch Surg 1987; 122:802-806. Alexander HR, Candela FC, Dershaw D, Kinne DK. Needle-localized mammographic lesions. Arch Surg 1990; 125:14411444.
180:
403-407.
750
#{149} Radiology
December
1991
Kambiz
MD
#{149} M.
Nonpalpable of Stereotaxic Fine-Needle
Lisa
terms:
Biopsies, #{149} Breast
biopsy, 00.125 00.31, 00.32
Radiology
I
From
1991;
(L.F.K.),
technology neoplasms,
and
Breast, diagnosis, #{149}
181:745-750
the Departments
(K.D.), Pathology
of General
(M.L.Y.),
Radiology
Surgery
Cytopathobogy
(P.M.J.),
Rush-Presbyte-
nan-St Luke’s Medical Center, 1653 W Congress Pkwy, Jelke Bldg. Rm 769, Chicago, IL 60612. From the 1989 RSNA scientific assembly. Received revision
dress
March 8, 1991; received July
reprint
requests
2 Current address: Biology, University
0 RSNA,
revision requested 8; accepted July
D
May 7; 23. Ad-
of Molecular Chicago.
F. Kluskens,
and
Reynolds
ESPITE
MATERIALS From
June
mographically for
1991
See also the editorial byjackson (pp 633-634) in this issue.
#{149} Larry
advances in mammographic technique, it remains difficult to correctly predict the benign on malignant nature of mammographically detected breast lesions. Even with these improvements, most lesions detected nadiographically and recommended for biopsy are benign at subsequent histologic examination. The positive predictive value of mammography in recent reports ranges from 14% to 38% (1-10). In 1986, steneotaxic fine-needle-aspiration (FNA) cytologic study of occult breast lesions was introduced into the United States in an attempt to overcome this pnoblem (11). Results from this and other studies showed that the procedure has a false-negative rate of 5%-14% and an inadequate sampling rate of 10%-26% (12-15). Also, most pathobogists in the United States have bess experience in breast cytology than in histology, and cytologic diagnoses are received by clinicians with less confidence. In 1988, we added fine-needle conetissue biopsies to FNA cytologic assessment in our diagnostic evaluation of nonpalpable, mammographically detected breast lesions to determine whether an acceptable and safe sobution could be found to reduce the rate of unnecessary open-breast biopsies. A mechanized fine-needle cone biopsy device (Biopty; Bard Unobogicab, Coyington, Ga) that has been used under sonographic control by physicians to procure tissue samples of suspicious liven, chest, and prostate nodules (1619) was used with our stereotaxic approach. This report presents the results of our experience in 250 such cases.
to K.D. Department of Chicago,
MD2
MD, PhD
#{149} Peter
M. Jokich,
MD
Breast Lesions: Findings Needle-Core Biopsy and Aspiration Cytology’
Two hundred fifty mammographically detected nonpalpable breast lesions suspicious for malignancy in women who underwent routine screening mammography were stereotaxically localized. Fine-needleaspiration (FNA) cytologic specimens and needle-core biopsy specimens were obtained before open biopsy in every case. Seventy-six lesions (30.4%) were malignanL Sixty-three (83%) of these 76 cancers were 1 cm long or smaller. Needle-core biopsy alone was used to diagnose conclusively 41% (ii = 31) of these cancers, while FNA cytologic study alone was used to diagnose 32% (n = 24). No false-positive results occurred with either test. The same diagnosis was reached in 54% (n = 41) when the combined results of both needle tests were considered. In applying the two needle tests to 125 mammographically defined low-suspicion lesions, 85 (68%) were found to be benign by means of either one or both needle tests; there was one lobular carcinoma in situ. By applying this algorithm, 85 (34%) of 250 patients with abnormal mammograms, or one-third of all patients recommended for open biopsy, might have avoided surgery. Index
Yaremko,
Imaging
carcinoma
AND
METHODS
1988 to May 1990, detected lesions in 242
women
250 mamsuspicious
study
was
granted by the Institutional Reand informed consent was obtained from each patient. For the punposes of this study, every abnormal mammogram depicting either soft-tissue masses or microcalcifications was reviewed by two investigators (K.D., P.M.J.) and was classified as having a high, intermediate, or low degree of suspicion for malignancy as follows: (a) A high index was indicated by a solid soft-tissue mass having highly irregular or stebbate borders; fine-needle-like and linear branching microcalcifications, either clustered or scattered; extremely polymorphic clustered microcalcifications; or an irregularly marginated solid mass containing suspicious or indeterminate microcalcifications. (b) An intermediate index was indicated by a solid soft-tissue mass with partially irregular or obscured borders or indeterminate microcalcifications. (c) A bow index
view Board,
was
indicated
by
a 0.5-1.5-cm
well-defined
mass, a 1-2-cm asymmetric clustered calcifications that
opacity, and were primarily rounded and rather uniform in size. If comparison mammograms were available, lesions were moved up or down on the scale of suspicion, depending on whether they were new or stable for 2-3 years; the latter group would be followed by means of mammography. The size limit of lesions for needle biopsy was lowered from 1 cm to 0.5 cm with experience. During the batter
part
of the
of patients with for malignancy
period,
approximately
mammograms
20%
suspicious
underwent
additional
di-
agnostic mammography consisting of spot-compression magnification views and/or ultrasound. The details of stereotaxic localization and FNA cytologic studies have been previousby
described
(12).
To be
lesion was stereotaxically the mammographic unit Fischer
Imaging,
Denver).
brief,
each
revisualized (Mammotest; The
skin
on
over
the puncture site was anesthetized with 1 % lidocaine, and a double-lumen coaxial needle (Cook Urobogical, Spencer, md) was inserted into the breast. The accuracy of needle localization was verified with stereoradiographs obtained before samplings. At least two and sometimes three
undergoing
routine screening mammography prospectively studied. Permission
were for this
Abbreviation:
FNA
=
fine-needle
aspiration.
passes for cytologic assessment were made with the inner 20-gauge Franzen-type needle. The core biopsy procedure performed with a 20-gauge Biopty needle fobbowed immediately after the cytologic aspiration. Again, two and sometimes three samples were obtained through the same puncture sites. Stereoradiographs of the needle in the lesion were obtained to document the part of the lesion obtained for biopsy.
In September
1989,
after
130
(120
cases).
gun had a lighter needle penetration, 17 mm.
The
The
spring, and
result
was
1
Histologic
to Radiographic
Correlation
Mammographically
Suspicious
Appearance
Nonpalpabbe
Suspicion
Index
of
and Type of Lesion
High or intermediate Mass
(n
=
No.
49(20) 76 (30)
10(6) 50 (29)
39(51) 26 (34)
80 34
93
84
9 (12)
9
(37)
32 (13) 250(100) in parentheses
a substantial
confirm the location of the wire. Only in 10% of cases were orthogonal views obtamed
for
following
(a) Two
intravenous
were
FNA
cytologic
biopsy by two
studies;
M.L.Y.,
without
knowledge
biopsy
Open-biopsy
to the Surgical where they were and processed for study.
fresh
specimens
needle-core
and
nee-
biopsies)
of the results
in a weekly
to standard
al-
specimens were reviewed of us (L.F.K., FNA cytologic
review
ries, and radiographs tobogic specimens ing
95%
of open
of cases,
histo-
by the authors. Cywere evaluated accord-
cytologic
criteria
(20,21)
and classed as negative (benign epitheliab elements only), atypical, suspicious for carcinoma, positive for carcinoma, or insufficient for diagnosis. The criterion for adequacy of cytologic specimens was more than six epithelial clusters of more than 10 cells each. Needle-core biopsy specimens were evaluated for size (representative of the size of a needle core, or approximately 7 mm bong) and presence of breast epithebial elements and stromal tissue.
The
epitheliab
elements
were
as-
sessed as benign, hyperplastic with or without atypia, suspicious, or malignant according to standard histopathobogic criteria (22). A needle-core sample was classed as insufficient if the histologic specimen was less than 7 mm long (one piece or total length of several pieces) or extensively fragmented, or if crush artifact precluded visualization of intact epitheliab cells. Finally, the findings of open biopsy were reviewed and results of the three modalities
compared.
After cytologic and needle-core biopsy specimens of the suspicious lesion were obtained, the location of the lesion in the breast was doubly marked with 0.3 mL of methylene blue dye injected through the localizing needle. A hook wire was inserted through the same needle. In 90% of cases, stereoradiographs were obtained to
746
#{149} Radiology
was
lesions
reasons:
Department.
with
separate
the
specimens
All
(%)
(48)
30 (17) 174(100)
2 (3) 76(100)
6 30
im-
Pathology
dle-core blindly
Value
are percentages.
cohob, and Papanicolaou staining. The needle-core biopsy specimens were placed in 50% Bouin fixative solution and submitted for routine histopathobogic processing and evaluation by members of the Surgical
Pathology Department examined, sectioned, routine histopathobogic
Predictive
faster speed of sampling notch of
fixation
sent
Positive
Biopsy Malignant
been localized in the same breast, and (b) the accuracy of localization was in doubt (eg, small lesions in a dense breast or a far posterior location). Patients were then taken to an operating room for excision of the lesion with the use of local anesthesia with or without
processing,
of Open
of
125)
Microcabcification Overall Note-Numbers
for 250
Benign
Mass
Monopty
of Suspicion
Lesions
Lesions
Microcalcification Low(n=125)
gun towas this
provement in the specimen quality and adequacy rate. Cytologic specimens were sent immediately to the cytopathology laboratory for cytospin
and Index
Breast Results
Mammographic
proce-
dures were performed, a new biopsy (Monopty; Bard Urobogical) that was tally disposable became available and exclusively used for the remainder of investigation
Table
had
sedation.
infrequently
General
anesthesia
given.
RESULTS Two
fifty lesions in 242 examined. The average age of the patients in this study was 53 (range, 30-85) years. Seventy-six malignant (30.4%) and 174 benign (69.6%) lesions were verified with open biopsy, for a positive predictive value of 30.4% for mammography in this series. Fifty-four cancers (71%) were invasive, and 22 (29%) were in situ. Overall, 108 lesions (43.2%) were suspicious microcalcifications, of which 29 (26.9%) were histologically proved to be malignant. One hundred forty-two lesions (56.8%) were soft-tissue abnormalities, of which 47 (33%) were malignant. The distnibution of lesions in the three mammographic categories and subsequent histologic findings are shown in Table
women
1. For
hundred
were
practical
reasons and simplicity mammognaphic lesions of high and intermediate suspicion were grouped together. This arbitrary division shows that it is possible to separate abnormal mammograms into two groups with a highly significant (x2 test; P < .001) difference in positive predictive value for malignancy. Half of all lesions (ii = 125), however, were in the low-risk mammographic category, with a prac-
of presentation,
ticab management outcome for needle tests. Table 2 shows the results of FNA cytologic assessment, needle-core biopsy, and both tests against the re-
sults of open biopsy. As expected, greater number of definitive diagnoses, either benign or malignant,
a
were
reached with needle-core biopsy than with FNA cytologic assessment. Twenty-four (32%) of 76 malignant lesions were conclusively diagnosed with FNA cytologic study, and 31 (41 %) were conclusively diagnosed with core biopsy. There was no falsepositive result. The sensitivity and specificity of these two tests were determined by including atypical, suspicious, and malignant lesions as positive and benign lesions as negative. Insufficient biopsy material for histologic diagnosis was regarded as no test. Thus, the sensitivity was 86%, specificity was 72%, and rate of insufficient results for FNA cytologic assessment was 24%. The same results for
core biopsy were 71%, 96%, and respectively. When the Biopty needle was replaced with the technically improved Monopty device after the first 130 cases, the rate of insufficient tissue sampling was reduced from24% to 8%. Table 2 also shows the combined results of FNA cytologic study and needle-core biopsy in 250 lesions. The 17%,
following nancy,
were
and chosen:
criteria for indeterminate Benign
benignity, maligdiagnoses lesions
were
those for which either both needle tests were negative or one was negative and the other was deemed insufficient. Atypical lesions were those for which either cytologic study or core biopsy showed dysplastic cells or atypical hyperplasia. Suspicious besions were those for which either one or both tests showed cells or tissue suggestive of but not definitive for malignancy. Malignant lesions were those for which either one or both needle lignancy.
tests revealed Employing
(54%) of 76 malignant conclusively diagnosed
evidence of masuch criteria, 41 lesions with
were a sensi-
December
1991
Table 2 Correlation
of Results
of FNA Cytologic Assessme Suspicious Breast Lesions
250 Mammograph.ically
nt,
Needle
-Core Biops y, and
Both
Techniques Open
Needle
Findings
(n
Cyto = 250)
Benign Atypical
99 41
Suspicious
25
Malignant Insufficient Note-Both
(n
both cytologic
155 11 10 31 43
assessment
and
needle-core
Table 3 Mammographic and Pathologic Features Cytologic Assessment and Needle-Core Lesion No.
Mammographic Appearance
-
Both = 250)
Cyto (n
126 39 27 41 17 biopsy,
of Malignant Biopsy
(n
=
biopsy
Missed
I
IDC
-
-
-
DCIS DCIS
-
-
IDC
I
-
Microcalcifications
I
-
-
-
8
Microcalcifications Microcalcifications
-
-
DCIS DCIS DCIS DCIS
9
Microcalcifications
I
I
DCIS
irregular
9-mm 5-mm
irregular mass stellate mass
4 5
Microcalcifications Microcalcifications
6
7
=
mass
findings, I = insufficient carcinoma in situ, IDC
Cyto
Correlation
of Combined Classification
FNA Cytologic
Assessment
in 125 Abnormal
Mammograms
and Needle-Core
Biopsy
Findings
Benign
with Lesions
of Low-Suspicion (n
FNA Cytologic and Core Biopsy (n = 125)
Benign (n = 115)
=
Classification 125) Malignant
(n
10)
=
85 20
85
0
18
2
Suspicious
8
Malignant Insufficient
6 6
6 0 6
2 6 0
Atypical
for diagnosis
tivity of 89%, specificity of 75%, and insufficiency rate of 7%. Sixty-three (83%) of 76 cancers were 1 cm or smaller in diameter at mammography. None of the 13 cancens banger than 1 cm was found to be benign by means of either needle test. Compared with FNA cytologic examination, needle-cone biopsy abbowed a superior (26 vs 15) diagnosis of malignancy when a mass was seen on the mammogram. On the other hand, FNA cytologic testing scored better (nine vs six) in the diagnosis malignant microcabciflcations. Also, (45%) of 29 malignant micnocabciflca-
of 13
cytologic
(n
9 15 16 24 12 assessment
were ductab nine lesions ameter.
=
Both
76)
(n
19 7 8 31 11
=
76)
8 10 16 41 1
only.
carcinoma were 1 cm
in situ. on less
All in di-
Application opsy method
Histologic Combined Assessment
=
Core
Cyto = 76)
(n
DISCUSSION
tissue for diagnosis. = invasive ductal carcinoma.
Table 4 Histologic
Both = 174)
Table 4 shows the results of the combined needle tests applied to lowrisk lesions. The insufficiency rate was 4.8% (six of 125) with no false-negative results but one case of lobular carcinoma in situ. The sensitivity was 100%, specificity was 78%, and negative predictive value was 100%.
Histologic Assessmentt
I
6-mm
(n
118 29 11 0 16
Both
with
Core Biopsy Findings*
only,
-
1
174)
136 4 2 0 32
needle-core
Lesions
=
in
Malignant
Core
174)
=
90 26 9 0 49
Core
Cytologic Assessment*
2 3
negative t DCIS = ductal *
(n
B iopsy
Biopsy
Benign
Core = 250)
24 61
for diagnosis =
Biopsy
Resul ts of Open
with
tions were falsely reported as benign with needle-core biopsy; only three (10%) of 29 were falsely reported as benign with FNA cytologic assessment. Histologically proved fibnoadenomas presented a sampling problem: These lesions shifted by about 5 mm as judged on postinsention radiographs. Table 3 presents the details of nine lesions missed with both cytologic examination and needle-cone biopsy. Six of the nine were micnocalciflcations. Only one of six lesions first seen as microcalcifications was invasive ductal carcinoma; the remaining five
of the needle-core biin conjunction with the steneotaxic device (Mammotest) brings a new dimension to the diagnosis of nonpalpabbe breast lesions suspicious for malignancy. The current average ratio of benign to malignant lesions for open biopsy of suspicious mammographic lesions in the United States is 3:1 (1,4,9). This is partly due to the unfavorable medicolegal risk exposure in this country beading to increased frequency of biopsy of bow-suspicion lesions. Neverthebess, the likelihood of abnormal mammograms being predictive of malignancy ranges from 10% in asymmetric panenchymal opacities and moderately suspicious microcalcifications to 50%-70% for stelbate masses and branching microcalcifications (5). The lesions of the first category in the “lighter shade of gray” may be safely followed with interval mammography, as recently reported in a webbconducted study (10). The number of unnecessary excisional breast biopsies may also be reduced by the application of stereotaxic fine-needle biopsy in conjunction with mammography. Azavedo et al reported only one cancen missed in 2,005 lesions examined in this way (23). Other investigators, also from Europe, reported favorably on their experience with steneotaxic fine-needle biopsy (14,24). Our results Ra1inlnov
#{149} 747
b.
a. Figure 1. magnification,
same
size
(a) Microscopic
appearance of a 1.5-cm fibroadenoma x 40.) (b) Microscopic appearance of a 1 .4-cm as that in a. (Original magnification, x40.)
show that stereotaxicably guided needle-cone biopsy, either alone on in conjunction with FNA cytologic assessment, is not a substitute for open biopsy in every instance. If selectively applied to mammographicalby defined low-risk lesions in which the prior probability of malignancy is less than 10% (Table 1), however, an excelbent degree of diagnostic reliability (100% sensitivity and negative predictive value) may be achieved (Table 4). We found the most serious limitations in the method to be (a) target size and mobility and (b) technique.
obtained by means invasive ductal carcinoma
ABNORMAL
Size
and
Mobility
(25).
All nine
ble
false-negative
cases
748
#{149} Radiology
needle. (Original a needle the
(n
INTERMEDIATE-
& 111CR-SUSPICION = 125)
LESIONS
UNif COST
t
,i
RE
BIOPSY
$11100
.,..5,
. NEEDLE
.
LOCALI
m
± NEEDLE
. . -.
:
5’..
$100 ‘
MAMMOGRAPHY (6.12 MONTHS)
BIOPSY
(LUMPFrrOMt
$2,100
I $102,000
$152,100
I
GRAND TOTAL
j
AvERAGE
$254,000 NEEDLE
I OPEN
BIOPSY
COST
P
CASE
Low-suspicion IntermediateFigure lesions.
EXCISI( 15
+ .#{149} OPEN
!
(Ta-
3) occurred among lesions first seen as either microcalcifications on a soft-tissue mass smaller than 1 cm, suggesting that mammographicably highor intermediate-suspicion besions of this size should undergo open biopsy. Also, seven of nine missed lesions were in situ carcinomas that might have shown a change in appearance at interval mammography if they had not been excised. Needle-cone biopsy can definitely be used to diagnose invasive carcinoma (Fig ib), but it is unlikely to allow confirmation of ductab carcinoma in situ. Suspicious microcabciflcations, which constituted 43.2% (n = 108) of all lesions in our series and 38% (n = 29) of malignant lesions, presented a special problem. This subgroup of breast lesions represents the earliest form of detectable breast cancer. Malignant micnocabcifications represent islands of cancer (ducts on lobules) less than 1 mm in diameter and therefore equal to the tip of cytology
urr COST
,!‘ .
r1i:
TOTAL
Seventy-eight percent of all lesions and 83% of cancers were equal to on less than 1 cm in diameter, compared with 40% reported a decade earlier
.. ,
Monopty biopsy with
MAMMOGRAMS
.
LSSIONS 123)
TOTAL SUB.
Target
of core biopsy with a 20-gauge obtained by means of core
2. Proposed Related costs
mab findings,
insuff
algorithm for each =
insufficient
mammogram & high-suspicion
$2,000
$3,000
mammogram
for management of nonpalpabbe diagnostic alternative are in the
mammographically shaded areas. Abnorm
depicted
abnor-
=
findings.
and cone biopsy needles. It is of interest that FNA cytologic assessment scored better (nine malignant and six suspicious lesions) than the needlecore biopsy (six malignant lesions and one suspicious lesion) in 29 malignant microcalcifications. We speculate that the reason for this difference is the to-and-fro motion effect of the Franzen needle that breaks down the supportive structures and frees the cells that are aspirated into the needle. We believe that needle-core biopsy and FNA cytologic assessment complement each other in the diagnosis of suspicious microcalcifications. Our experience shows that the correct diagnosis of these minute lesions is more likely with multiple passes of the needle (both types) into the besion, which we practice and strongly
recommend.
If the
suspicion for spite adequate
cancer remains high negative cytologic
cone biopsy biopsy
samples,
should
Adequate
of the
antee
fixation
about
to be
open immonot
lesion
guan-
when
by
the
it is needle.
in the location of the lesion suggestive of a fibroade-
as opposed is relatively
during
does
penetrated
which thebess,
an and
breast
of the
noma,
negative target
then
deand
be performed. compression
bilization
The shift is highly
mammographic
the
lesion
to a carcinoma, immobile. Nevershift
tissue sampling. may occur more a core
biopsy
results
in false-
Missing frequently sampling
the be-
cause the needle is mechanically fined into the lesion and the “feel” of tissue resistance experienced by means of December
1991
manual
penetration
logic
during
assessment
Technical
FNA
cyto-
is lost.
Aspect
There was a substantial improvement in the specimen quality noted by our
pathologist
when
the
fragmentation
as webb
surgical
Biopty
needle was replaced with the Monopty device midway through study. The prevalence of specimen as the
the
prohibiting
multiple
micro-
sampling
of
the target. Also, a small incision is necessary for skin penetration. The final choice of the needle may depend on the lesion size: the bigger needles for 1-2-cm soft-tissue lesions and the finer needles for microcalcifications. Diagnostic Mammograms
Plan
for Abnormal
Mindful of the current increasing liability exposure in the United States, we propose the following algorithm. We strive to reduce the chance of missing
breast same
the
diagnosis
cancer as the
of an
to 1%-2%, reported
early
which
is the
prevalence
for
open biopsy (6,27-29). The threshold of this risk acceptance may be raised with experience on in individual patients who refuse to undergo open biopsy. The algorithm is summarized in a flow chart Mammographically
mediate-risk ommended ization and FNA
cytologic
(Fig
2). high-
on inter-
lesions are generally necto undergo needle localopen biopsy. Optional assessment
performed
at the time of localization with immediate reporting (within 1 hour) serves two purposes: (a) A positive diagnosis of malignancy will assist the surgeon in performing a complete bumpectomy with tumor-free margins, thus avoiding the risk of a difficult and unsatisfactory ne-excisionab operation. This is particularly important in the present-day practice of breast consenvation. (b) In cases of two mammoVc
me 181
#{149} Number
3
management.
An
asymptom-
atic woman whose screening mammogram reveals a low-risk lesion is advised to undergo stereotaxic localization,
scopic crush effect decreased markedly, beading to a sample adequacy rate of 92% for Monopty compared with 76% for Biopty. The cone biopsy specimens procured with a 20-gauge needle (Fig 1) were generally adequate for histologic interpretation. Parker et ab (26) reported that a more satisfactory specimen sample was obtamed with 14-gauge Biopty needles. After completion of this study, we began using 18-gauge needles, as nonprotocol pathologists in our institution prefer to work with larger biopsy specimens. We believe that larger needles cause more bleeding, thus
graphically separate, yet suspicious, lesions in the same breast on extensive suspicious microcalcifications that exdude breast conservation, positive needle diagnosis of lesions will help the surgeon and the patient decide on
FNA
cytologic
assessment,
to undergo
mography within such an algorithm to the 125 bow-risk 85 lesions
would
interval
have
been
Shott,
versus
2.
is given
in Figure
2. Although
the
biopsy
for
250
cases
6.
advice;
Roshni
Patel,
AL, Schwartz AS. Clinically
GF, Seig SA, Paoccult breast be-
sions: localization and significance. Radiology 1987; 162:167-170. Tinnemans JGM, Wobbes TH, Holland R, et
Rowen
JP, Bassett
55, Simon
IS, Brown
DS.
biopsy for mammoin 561 patients. Can
Silverstein MJ, Gamagami P, Rosser RJ, et al. Hooked-wire directed breast biopsy and overpenetrated mammography. Cancer 1987; 59:715-722. Ciatto 5, Cataliotti
lesions
Z, Distante
detected
V.
Nonpal-
with mammogra-
phy: review of 512 consecutive ology 1987; 165:99-102. Hall FM, Storella JM, Sibverstone
cases.
Radi-
DZ,
Wy-
shak G. Nonpalpable breast lesions: recommendations for biopsy based on suspi-
7.
cion of carcinoma at mammography. Radiology 1988; 167:353-358. Hermann G, Janus C, Schwartz IS, Krivisky
B, Bier 5, RabinowitzjG. breast lesions: accuracy
8.
diagnosis.
Nonpalpable of pre-biopsy Radiology
mam-
mographic 323-326. Papatestas Tsevdos
AE, Herman C, Lesnick G.
D, Hermann G, Surgery for non-
palpable
breast
Arch Surg 1990;
lesions.
1987;
165:
125:399-402.
9.
of ab-
Acknowledgments: We gratefully appreciate the collaboration of Steven D. Bines, MD, Daniel J. Deziel, MD, Alexander Doolas, MD, Steven G.
to
Susan
BA, and Debra Babich, MD, assistance; and Jan Nunnalby help.
Needle-guided breast graphic abnormalities Surg 1986; 29:287-288.
cost
normal mammograms with that for the same number if handled according to the proposed algorithm. This cost savings is attained with no risk to patients. Also, one-third of women derive superior cosmetic results, and their unscarred breasts could be evaluated more satisfactorily at subsequent examinations. On the basis of the result of this and other studies, we believe that steneotaxic needle biopsy of nonpalpable breast lesions will play a significant robe in the future diagnosis and management of breast cancer. It will reduce the number of unnecessary breast biopsies and the associated cost and will encourage more women to participate in screening mammographic surveillance for early detection of breast cancer. U
for statistical
Rosenberg tachefsky
pable
of individual needle biopsy may be one-third that of excisional biopsy ($1,100 vs $3,000), the average cost for the low-risk lesions with associated open biopsies in 40 of 125 cases would raise the average diagnostic cost of this group to $2,000 versus $3,000 for high- and intermediate-risk lesions. Nevertheless, a 15% cost reduction is achieved if we compare the cost of excisional
PhD,
patients
165:523-529.
3.
4.
cost for stereoversus that of outat our institution groups of lesions
their
We also thank
al Mammographic and histopathologic correlation of nonpalpable lesions of the breast and the reliability of the frozen section diagnosis. Surg Gynecol Obstet 1987;
Benefit
The approximate taxic needle biopsy patient open biopsy in 1991 for the two
encouraged
in this study.
References
5.
Cost
who
BA,Julee Bangert, for their technical for her secretarial
diag-
of either
MD,
participate
mam-
6-12 months. If had been applied cases in this series,
nosed as benign by means one on both needle tests.
Wool,
1.
and cone biopsy. If both samples are reported inadequate or one is abnonmab (atypia, suspicious, on frankly malignant), open biopsy should be penformed. If one or both needle samples are adequate and benign, the patient is advised
Economou, MD, Keith W. Millikan, MD, Theodore J. Saclarides, MD, Edgar D. Staren, MD, PhD, Thomas R. Witt, MD, and Norman L.
10.
Franceschi
D, Crowe
J, Zollinger
R, et al.
Biopsy of the breast for mammographically detected lesions. Surg Gynecol Obstet 1990; 171:449-455. Sickles EA. Periodic mammographic
bow-up
of probably in 3,184 consecutive
benign cases.
fol-
lesions: results Radiology 1991;
179:463-468.
11.
Dowlatshahi
12.
Bibbo M, Dawson PJ. Cytologic diagnosis of occult breast lesions using stereotaxic needle aspiration. Arch Surg 1987; 122: 1343-1346. Dowbatshahi K, Gent HJ, Schmidt R, Jokich
K, Jokich
PM, Schmidt
PM, Bibbo M, Sprenger
E.
R,
Nonpalpable
breast tumors: diagnosis with stereotaxic localization and fine-needle aspiration.
13.
diobogy 1989; 170:427-433. Svane G, Silfversw#{228}rdC. dle biopsy of nonpalpable
Stereotaxic neebreast lesions:
cytologic and histologic findings. Acta diol [Diagnj(Stockh) 1983; 24:283-288.
14.
Lofgren
M, Andersson
Stereotactic
logic diagnosis sions. AJR
15.
Jackson
I, Lindholm
fine-needle
Ra-
aspiration
of nonpalpable
Ra-
K. for cyto-
breast
le-
1990; 154:1191-1195.
VP, Bassett
LW.
Stereotactic
fine-
needle aspiration biopsy for nonpalpable breast lesions. AIR 1990; 154:1196-1197. 16.
Anderssen
T, Erikson
B, Lindgren
PG, lvi-
lander E, Obeng K. Percutaneous ultrasonography-guided cutting biopsy from liver metastases of endocrine gastrointestinal tumors.
Ann
Surg
1987;
206:728-732.
Radiobosrv
#{149} 749
17.
18.
Goralnik CH, O’Connell DM, EL Yousef SJ, Haaga JR. CT-guided cutting needle biopsies of selected chest lesions. AJR 1988;
22.
151:903-907.
23.
Ragde
H, Aldape
trasound-guided gun superior
Urology
1988;
19.
32:503-506. Kaye KW, Horwitz
20.
sound-guided prostate biopsies using new automatic gun: analysis of 100 consecutive cases. J Endourol 1989; 3:155-162. Kline TS. Masquerades of malignancy: a
21.
review of 4,241 aspirates from the breast. Acta Cytol (Baltimore) 1981; 25:263-266. LinskJA, Franzen S. Clinical aspiration cytology. 1989.
2nd
ed.
24.
CA.
Transrectal
Philadelphia:
ultra-
Lippincott,
York:
histopa-
27.
Churchill-
G, Auer in 2594
G. Stereotactic mammographi-
cally detected non-palpable lesions. Lancet 1989; 1:1033-1036. Gent HJ, Sprenger E, Dowbatshahi K. Steneedle
cal diagnosis
26.
Diagnostic New
1987.
Azavedo E, Svane fine needle biopsy
reotactic
25.
J.
of the breast.
Livingstone,
HC, Bagley CMJr. UIprostate biopsy: Biopty
to aspiration.
Page DL, Anderson thobogy
localization
of occult
breast
and
28.
cytobogi-
lesions.
Ann
Surg 1986; 204:580-584. Moskowitz M. Significance of minimal breast cancers detected on screening. AIR 1981; 136:735-738. Parker SH, Lovin JD, Jobe WE, Burke BJ,
Hopper KD, Yakes WF. Nonpalpabbe breast lesions: stereotactic automated large-core biopsies. Radiology 1991;
29.
Landercasper J, Gundersen SB Jr, Gunderson AL, Cogbill TH, Travebli R, Strutt P. Needle localization and biopsy of nonpal-
pable lesions of the breast. Surg Gynecol Obstet 1987; 164:399-403. TinnemansjGM, Wobbes T, Hendriks JHCL, Van der Sluis RF, Luberes EJC, deBoer l-IHM. Localization and excision of nonpabpable breast lesions: a surgical evaluation of three methods. Arch Surg 1987; 122:802-806. Alexander HR, Candela FC, Dershaw D, Kinne DK. Needle-localized mammographic lesions. Arch Surg 1990; 125:14411444.
180:
403-407.
750
#{149} Radiology
December
1991
