Hind S. Teixidor, MD Charles A. Santos-Buch,

#{149} David

MD

Fine-Needle Specimens: and Cytologic

A. Wojtasek, #{149} C. Richard

MD Minick,

#{149} Ellen

Aspiration Correlation Findings’

The accuracy of fine-needle aspiration (FNA) cytologic diagnosis of nonpalpable breast lesions and the prevalence of neoplasm occurring in areas unrelated to the radiologic ab-

T

M. Reiches,

BS, MT (ASCP)

MD

ofBreast Biopsy of Histologic

The results of cytologic examination of the aspirates were compared with the results of histologic examination of the surgically excised tissue. This in vitro study may provide the optimal results that can be expected from in vivo FNA of nonpalpable breast le-

was performed in 101 surgically excised breast specimens. The exact area of the mammographic abnormality was aspirated with radiographic

use of screening mammography has increased the rate of detection of nonpahpable breast abnormalities. Abnormalities that are suspicious for neoplasm or are nonspecific often must be needle localized and undergo biopsy for definitive diagnosis. These procedures have proved to be costly, cosmetically disfiguring to the breast, and psychologically traumatic to patients. Only 15%-

control.

30%

Needle localization followed by surgical excisional biopsy of nonpalpable breast lesions was performed in 101 patients re-

normality guided

FNA

were studied. cytologic

Despite

Templateexamination

accurate

placement

of the needle for aspiration, seven of 101 aspirates (7%) yielded insufficient cytologic material. Ninety-four of the 101 aspirates (93%) were adequate for diagnosis. The cytologic diagnosis was benign in 58 (62%), atypical in seven (7%), suspicious for malignancy in four (4%), and malignant in 25 (27%). All cases diagnosed as suspicious or malignant and cases diagnosed as benign logic examination proved

five

of 58 at cytoto be ma-

lignant at histologic examination. In three of these five the malignancy was in the area of the radiologic abnormality; in two it was not. FNA cytologic

examination

in evaluating sions,

but

nonpalpable it is not

tologic examination cised lesions. Index terms:

be

helpful

breast

as accurate

of surgically

le-

as his-

ex-

Breast,

Breast, cytology, localization, 00.125, Radiology

can

1992;

biopsy, 00.1261, 00.1262 00.1262 #{149} Breast neoplasms, 00.30

184:55-58

From the Departments of Radiology (H.S.T., D.A.W.) and Pathology (E.M.R., C.A.S.B., C.R.M.), New York Hospital-Cornell Medical Center, 525 E 68th St, New York, NY 10021. From the 1991 RSNA scientific assembly. Received November 14, 1991; revision requested 1

December

26; revision received and accepted Address reprint requests to

February 24, 1992. H.S.T. 0 RSNA, 1992

HE

of the

lesions

subjected

to biopsy

are malignant (1-3); the rest are benign or atypical. It is desirable to decrease the number of surgical biopsies, particularly in cases of benign breast disease, provided the detection rate of cancer remains unaffected. Fine-needle aspiration (FNA) cytologic examination of nonpalpable breast lesions has shown encouraging results in differentiating benign from malignant diseases (4-10). It may not only help avert surgery in many patients (4,7,10,11) but will substantially reduce the cost of diagnosing breast cancer (5,6). This is contingent on accurate

placement

of the

dle, adequate sampling and a low false-negative logic results. Using

various

biopsy

nee-

of the lesion, rate of cyto-

localizing

of sampling

eliminated

by

directly

error

ferred

AND

METHODS

because of an abnormal mammoThe mammograms had been obat various radiology facilities. The

gram.

tamed

patients were not consecutive but included all those who had undergone needie localization over a 10-month period by the two radiologists involved in this study

(H.S.T.,

D.A.W.).

with

Needle

localization

radiography

specimen

a Senographe

ical Systems,

were

500-T imager

Milwaukee).

and

performed The

(GE Medresected

specimens were radiographed under perforated plate with grid coordinates. The compression plate of an E-Z-Em bin specimen used (E-Z-Em,

radiography Westbury,

device NY).

a Du-

was

The exact area of abnormality was idenand aspirated through the hole in the plate overlying the lesion (Fig la). A rifled

butterfly

pressure

were through

10-15 covering

applied

needle with

and

a 20-mL

negative syringe

used.

The needle was inserted once a 4-mm hole in the plate, and passes were made at various angles a larger area of tissue. This re-

sulted in sampling of approximately of breast tissue, including the targeted sion. The specimen was radiographed

2 cm3 le-

again with the needle in place to confirm the location of the needle in relation to the abnormal was inked

area (Fig ib). The aspirated site before removal of the plate, and

the whole histologic

specimen was submitted for examination. The histopatholo-

gist blocked (designated

and examined the “S” area)

the inked separately.

area The

was

aspirating

area of abnormality in the specimen with radiographic

MATERIALS

21-gauge

methods

and equipment, some investigators have reported a false-negative rate of 3%-6.7% for the diagnosis of carcinoma (6,12). Insufficient material has been reported in 8.1%-26% of cases (4-10,13). The reasons for obtaining insufficient material are believed to be related to the accuracy of needle placement and type of breast tissue sampled (8,12). We have attempted to assess the accuracy and limitations of the FNA biopsy procedure and to determine the prevalence of neophasm occurring in areas unrelated to the radiologic abnormality. Faulty needle placement as a source

sions.

the

excised control.

Abbreviations: DCIS = ductal carcinoma situ, FNA = fine-needle aspiration, LCIS lar carcinoma in situ.

in

= lobu-

55

size

of the

mately size

S area

varied

from

1 cm3 to 2.5 cm3, of the

radiologic

was

radiologist

pirate

on

95% later

ethanol stained

slides

and

were

needle

as-

fixed

interpreted

the

purpose

slides

were

senior cytotechnologist attending cytopathologist histologic slides were pathologist

in a

and

tions of the S area resected specimen results (ie,

into

a cytostudy

reviewed

by a

(E.M.R.) and an (C.A.S.B.). The reviewed by a senior The

cytologic

and of the were later

of cytologic

classified

by of this

(C.R.M.).

histologic

isfactory

the

that

were

For

the cytologic

The

rest

in the

solution. The slides were with use of the Papanicolaou

pathologist.

were

The

processed

smeared

four

technique

of the

the

manner.

The

surgical

on

abnormality.

of the specimen usual

approxi-

depending

rest of the compared.

examinations

five groups:

specimen

results

examina-

showing

(a) unsatfour

or

and fibrous mastopathy. In the remaining four specimens, insufficient cellular sampling was attributed to the type of breast lesion. One specimen contained DCIS; the other three showed fibrocystic disease and sclerosing adenosis. Ninety-four (93%) of the 101 aspirates had satisfactory cytologic material. The cytologic interpretation was benign in 58 (62%), atypical in seven (7%), suspicious for carcinoma in four (4%), and malignant in 25 (27%) of these 94 aspirates. Of the 58 cases diagnosed as benign at cytologic examination, 55 (95%) showed benign disease in the S area at histologic examination. Two of the 55 showed in situ carcinoma outside the S area: One was a single focus of

DCIS,

and

the

LCIS.

The

radiologic

other,

a single

focus

abnormality

of

in

the latter two cases was a cluster of calcifications. Of the remaining three cases (of 58) diagnosed as benign at cytologic examination, one involved an infiltrating ductal carcinoma that appeared radiologically as a stellate distortion (Fig 3); the second, a DCIS that appeared mammographically as a cluster of calcifications; and the third, a single focus of LCIS. Seven cases were cytologically diagnosed as showing atypia. Two proved histologically to be carcinomas, one an infiltrating ductal carcinoma, and the other an LCIS (two foci). The remaining five showed atypia rather than malignancy at histologic examination, and therefore

fewer groups of epithelial cells and/or fewer than 100 epithelial cells); (b) benign (ie, including benign epithelial groups without atypia, sclerosing adenosis, hyalinized

fibroadenoma,

fibrous

mastopa-

thy, fat necrosis, and intramammary lymph node); (c) atypical (ie, atypical ductcell hyperplasia with or without intraductal papillomatosis); (d) suspicious for malignancy (ie, abnormal cells that are almost certainly malignant but either are too scanty or show incomplete criteria for an unequivocal diagnosis of malignancy); and (e) malignant. The mammographic abnormalities that led to the biopsy were all nonpalpable. They included microcalcifications in 61 cases (60%) and soft-tissue abnormalities in 40 (40%). In the first group, the calcifications were in a cluster 3-10 mm in diameter in 56 of 61 cases and were more diffuse in the remaining five. The soft-tissue abnormalities consisted of masses in 37 cases and of parenchymal distortion in three. Twenty-seven of the masses were discrete and ranged in diameter from 0.5 to 1.5 cm. Two

masses

remaining and ranged

contained

calcifications.

The

10 masses were stellate in shape in diameter from 0.4 cm to 1.6

cm.

a. Figure 1. (a) Radiograph of surgically resected breast biopsy specimen under a perforated plate with grid coordinates to localize the mammographic abnormality for which the biopsy was performed. Calcifications and the localizing wire can be seen in the specimen. (b) Radiograph

of same

specimen

as in a after

insertion

of a butterfly

needle

the calcifications. Aspiration for cytologic examination was performed tologic and cytologic diagnosis was hyalinized fibroadenoma.

through

through

the

hole

that

overlying

hole.

RESULTS The histologic diagnoses of the surgically resected specimens were cornpared with the cytologic results and mammographic findings (Table). The cytologic examination was considered unsatisfactory in seven (7%) of the 101 specimens due to insufficient material. In three of the seven, the specimen radiograph showed that the hesion was located at the periphery of the excised specimen (Fig 2), making it technically difficult to aspirate the area. This difficulty was considered the reason for obtaining insufficient material. At histologic examination one of these three specimens contamed a single focus of DCIS. The other two showed fibrosis, adenosis, 56 #{149} Radiology

July

His-

may be considered as false-positive this group. A total of 29 cases were diagnosed as either suspicious for malignancy (n = 4) or malignant (n = 25). Histologically, all 29 proved to be mahig-

nant,

with

situ

carcinomas.

positive

14 infiltrating

and

There

diagnoses

were

in this

in

15 in no

false-

group.

DISCUSSION

more

reliant

on the

by

number

Despite placement,

of unnecessary

biopsies.

FNA

biopsy may help clinicians reach such an objective (4,7,10) but is not yet widely practiced or accepted. The major difficulties of FNA biopsy are accurate needle placement and obtainment of sufficient cytologic material from breast lesions that have

As the number of nonpahpable breast lesions being found at mammography increases, clinicians are

becoming

of the radiologist for the management of these lesions. The low specificity of mammography, however, which ranges from 15% to 30% for the diagnosis of carcinoma (1-3), has led to an increasing number of surgical biopsies being performed for diagnostic purposes. Efforts are being made to increase the specificity of the mammographic diagnosis to reduce the

opinion

abundant reactive fibrosis. Many reports of the usefulness of FNA biopsy have appeared in the recent hitera-

ture,

but

different

techniques

authors count

equipment

have

been

(4-10,12,13), in part for the

results.

The

authors’

ing insufficient ranged

the

from

23-gauge

8.1%

used and

important

after

Figure 2. Radiograph of breast biopsy specimen under the localizing plate with butterfly needle in place. Calcifications are at the periphery of the specimen (arrow). The aspirate contained insufficient material for cytologic examination.

of obtain-

material

to 26%.

ranged did

various

not

The

from seem

size

of

21- to to be an

factor.

Therefore,

evaluate

rates

cytologic

needle

by

which may acvariation in their

we

FNA having

have

cytologic eliminated

attempted

to

examination or minimized

needle placement as a source of error for adequate sampling of the lesion. Concern has also been expressed as to whether a neoplasm in the vicinity of the radiologic missed with ficulties were

abnormality FNA biopsy addressed

might be (9). Both difin our study

placing

in the

the

area

biopsy

of the

flee-

radio-

logic abnormality in all cases, then submitting the entire surgical specimen for histologic examination after having inked the area that was aspirated, to allow the histopathologist to include it in the S area and to study it separately

from

the

rest

of the

speci-

men. the accuracy 7% of the

of needle lesions yielded

unsatisfactory cytologic samples. In 3%, this was attributed to technical difficulties because the lesion was hocated at the periphery of the specimen, thus not allowing adequate aspiration. The remaining unsatisfactory samples were attributed to the type of breast tissue aspirated. Abun-

dant

and

used

accurately

dle for FNA

fibrosis,

scarring,

and

lesions

with low cellularity or a small number of ductules reduced the chances of obtaining representative material (8,12). Similar lesions were successfully aspirated in the rest of our material, however, and therefore, other undetermined technical factors may have played a role. This rate of unsatisfactory samples would have decreased further had we aspirated the specimen more than once whenever insufficient material that could be seen with the naked eye was expelled from the needle. It has been shown that three or four aspirations of a breast lesion produce a better yield than does one (14). Moreover, having a cytopathohogist present at the time of needle aspiration to confirm the adequacy of the sample is helpful but not always possible. Although it is difficult to eliminate insufficient cytologic material as a sampling error, it can perhaps be reduced to 4% or less if the needle is accurately placed and if multiple needle passes are performed. Lesions riah should repeated

that yield insufficient matebe restudied, either with aspiration or with surgical

excision.

In our

insufficient

study,

material

cases were

involving considered

neither benign nor malignant and were not included in the statistical analysis

did

not

of the

provide

A specific

a.

b.

Figure 3. (a) Mediolateral-view mammogram of the breast with localizing needle and wire in place. The tip of the needle is close to a soft-tissue abnormality with stellate distortion (arrow). (b) Radiograph of the resected lesion in a, showing the tip of the butterfly needle in the area of abnormality (arrow). The cytologic diagnosis was benign. Histologic examination showed a 5-mm infiltrating ductal carcinoma. Volume

184 #{149} Number

1

results

cytologic

diagnosis

because

they

information. of benign

breast disease was made cytologically in 58 cases and was confirmed histologically in 55. In two of the 55 cases-the only two in the entire series in which malignancy was found unrelated to the radiologic abnormality-single foci of carcinoma in situ were found in the specimen outside the S area. These two cases were not considered to give false-negative re-

Radiology

#{149} 57

sults

because

they

did

not

lie in the

cytologically sampled area. Moreover, one of the two cases showed a single focus of LC!S. Foci of LCIS are mcidental findings not uncommonly found in tissue adjacent to, but not in the area of, the mammographic abnormality for which the biopsy was performed (9). Of the 58 cases diagnosed cytologically as benign, three showed carcinoma in the S area at histologic examination. These results were considered false-negative. A low false-negative rate of 3%-6.7% has been reported in in vivo studies (6,9,12). A review of the mammographic findings in cases reported as cytologically benign can play an important role in further evaluating such lesions. Those lesions appearing suspicious mammographically should be restudied with either a repeated FNA

or surgical

Of the diagnosed atypia,

tologic nosis

excision.

seven specimens that were cytologically as showing two showed carcinoma at hisexamination. A cytologic diag-

of atypia

is an

indication

for

sur-

gical biopsy (4,7,9,13). Therefore, carcinomas in this group should not be missed because all patients with atypia will undergo surgical resection. Investigators of in vivo studies have treated atypia similarly (7,8). All 29 specimens diagnosed cytologically

as malignant

or suspicious

for malignancy were found at histologic examination to be cancerous. Similar results have been reported in in vivo studies in which no false-positive results were found in those cases diagnosed cytologically as malignant or highly suspicious of malignancy (6-9). Of the 14 inifitrating cancers in this group, 11 (79%) appeared mammographically as a mass or parenchymal distortion. There were 15 in situ carcinomas, 14 (93%) of which showed only calcifications at mam-

therapeutic management. FNA cytologic assessment

3.

Therefore, may not

4.

provide sufficient information to pursue breast conservation treatment, in interested patients. In cases involving adequate cytologic material there were a total of 34 carcinomas, 31 (91 %) of which were correctly diagnosed with cytologic examination.

This

includes

cases

5.

6.

single was

focus

of LCIS,

probably

a finding

incidental

and

the

information

may

for diagnostic

Acknowledgment I. Henschke, tical analysis

help

accuracy

avert

purposes.

We are grateful PhD, MD, for her help of the data.

lesions:

prospective

11.

12.

13.

surgical

U to Claudia in the statis-

comparison

aspiration

cytology.

fine-needle

14.

Radiology

I, Lindholm

aspiration

K.

for cyto-

logic diagnosis of nonpalpable breast lesions. AJR 1990; 154:1191-1195. Hann L, Ducatman BS, Wang HH, Fein V, Mclntire JM. Nonpalpable breast lesions: evaluation by means of fine-needle aspiration cytology. Radiology 1989; 171:373-376. Helvie MA, Baker DE, Adler DD, Andersson I, Naylor B, Buckwalter KA. Radio-

graphically guided fine-needle of nonpalpable breast lesions.

is

DC.

with surgical biopsy results. AJR 1990; 155: 977-981. Ciatto S, Del Turco MR, Bravetti P. Nonpalpable breast lesions: stereotaxic fine-

Stereotactic

10.

Wiens

Ra-

stereotactic finecytology of nonpalpable

1989; 173:57-59. Lofgren M, Andersson

96.5% (84 of 87). If the seven cases of cytologic atypia are not excluded, however, the sensitivity is 91 % (31 of 34), the specificity is 92% (55 of 60), and the accuracy is 91% (86 of 94). The results of this study suggest that FNA cytologic examination of nonpahpable breast lesions is a valuable tool in improving the specificity of mammography. It should be used selectively in cases in which cytologic biopsy

JL, Trego

LL, DavisJR,

needle

9.

stereotaxic

Fajardo

radio8.

with

aspiration.

breast

7.

Lancet

HJ, Schmidt R,Jokich E. Nonpalpable

diagnosis

that

to the

overall

tumors:

lesions.

localization and fine-needle diology 1989; 170:427-433. Mammography-guided needle aspiration

a

logic abnormality. If the seven cases diagnosed as atypical at cytologic examination are considered neither benign nor malignant and are excluded from the statistical analysis, the sensitivity of FNA cytologic assessment in this series is 91 % (29 of 32), the specificity is 100% (55 of 55),

nonpalpable

1989; 1:1033-1036. Dowlatshahi K, Gent PM, Bibbo M, Sprenger

breast

from

mas were missed at cytologic examination. One of these three was only

Rosenberg AL, Schwartz GF, Feig SA, Patchefsky AS. Clinically occult breast lesions: localizations and significance. Radiology 1987; 162:167-170. MeyerJE, Kopans DB, Stomper PC, Lindfors KK. Occult breast abnormalities: percutaneous preoperative needle localization. Radiology 1984; 150:335-337. Azavedo E, Svane G, Auer G. Stereotactic fine-needle biopsy in 2,594 mammographi-

cally detected

each of the three cytologic groups of atypical, suspicious, and malignant specimens. Three (9%) of the carcino-

aspiration Radiology

1990; 174:657-661. FrisellJ, Eklund G, Nilsson R, Hellstr#{246}m L, Somell A. Additional value of fine-needle aspiration biopsy in a mammographic screening trial. BrJ Surg 1989; 76:840-843. Bibbo M, Scheiber M, Cajulis R, Keebler CM, Wied GL, Dowlatshahi K. Stereotaxic fIne needle aspiration cytology of clinically occult malignant and premalignant breast lesions. Acta Cytol 1988; 32:193-201. Evans WP, Cade SH. Needle localization and fine-needle aspiration biopsy of nonpalpable breast lesions with use of standard and stereotactic equipment. Radiology 1989; 173:53-56.

Pennes needle

DR. Naylor B, Rebner M. Fine aspiration biopsy of the breast: in-

fluence sample

of the number of passes and the size on the diagnostic yield. Acta

Cytol

1990; 34:673-676.

References 1.

58 #{149} Radiology

2.

mography. Five cases involving DCIS showed muhtifocal or extensive disease. Muhtifocality cannot be determined cytologically and may play an important role in determining the

Moskowitz M. Minimal redux. Radiol Clin North 113.

breast cancer, Am 1983; 21:93-

July

1992