Microcalcifications of non-palpable breastlesions detected by ultrasonography:correlation with mammography andhistopathology

C.-S. Huang*†, C.-Y. Wu‡, J.-S. Chu**, J.-H. Lin†, S.-M. Hsu** and K.-J. Chang*

Departments of *Surgery and **Pathology and †Division of Medical Ultrasound, National Taiwan UniversityHospital, National Taiwan University College of Medicine; ‡Department of Radiology, Cathay General Hospital,Taipei, Taiwan

Key words: NON-PALPABLE BREAST LESIONS, MICROCALCIFICATIONS, MAMMOGRAPHY, HISTOPATHOLOGY,

ULTRASONOGRAPHY, BREAST CANCER

ABSTRACT

Objectives Microcalcifications are generally not dem-onstrated well on ultrasonography. In this study, weattempted to demonstrate the usefulness of high-resolutionultrasonography in the detection of microcalcificationsassociated with non-palpable breast cancers.

Design Fourteen patients with non-palpable breast lesionsin whom microcalcifications were detected or suspected byultrasonography and one patient in whom microcalcifica-tions were detected on mammography only were includedin the study. Mammography and analysis of biopsy speci-mens were performed in each patient and the findings werecorrelated with the ultrasonographic findings. Ultrasono-graphy and mammography were performed independentlyby different physicians at different times.

Results In three patients ≤ 30 years of age, who were notat high risk of breast cancer and who had no evidence ofcancer on palpation, high-resolution ultrasonographyclearly showed microcalcifications but no mass. Two ofthese patients had ductal carcinoma in situ and one hadsmall invasive carcinoma with extensive comedocarci-noma. Among the other 12 patients with non-palpablebreast lesions, ultrasonography detected microcalcifica-tions accurately in six and suggested possible microcalcifi-cations in a further four. Microcalcifications in all of theseten patients were confirmed by mammography thereafter.Four of these ten patients had ductal carcinoma in situ,with or without invasive carcinoma. Of the remaining twopatients, one demonstrated false-positive findings and onefalse-negative findings on ultrasound. On high-resolutionultrasonography, microcalcifications produced the appear-

ance of twinkling stars (bright dots in different planes) in adark sky (contrasted against ill-defined hypoechoicpatches), corresponding on histopathology to groups ofexpanded ducts with increased cell density with or withoutnecrosis.

Conclusion High-resolution ultrasonography may beused for detection of microcalcifications in non-palpablebreast lesions. Ultrasonography is helpful in screening forearly breast cancers, especially in young patients who are atrisk for breast cancer and in whom mammography is notusually carried out.

INTRODUCTION

When there is a non-palpable cancer in a dense breast1–5,clustered microcalcifications are one of the primary X-raymammographic findings, and sometimes the only finding.Because difficulties exist in differentiating malignant frombenign calcifications5,6, biopsy aided by wire localizationwith careful pathological examination is usually needed forassessment of the nature of a breast lesion1–6. Ultrasono-graphy can also detect coarse calcifications with acousticshadowing in calcified fibroadenoma7. However, ultra-sound is generally believed to be unsuitable for the detec-tion of fine microcalcifications in non-palpable breastlesions8–15. A mass ultrasound screening of 18 539 womenfor breast cancer in Japan revealed 22 breast cancers, 16 ofwhich were early breast cancers and 13 of which werenon-palpable tumors14. This Japanese study did not includepatients with known microcalcifications detected by

Correspondence: Dr C.-S. Huang, Department of Surgery, National Taiwan University Hospital, 7 Chung-Shan S Road, Taipei, Taiwan

Ultrasound Obstet Gynecol 1999;13:431–436

ORIGINAL PAPER Received 21–1–98Revised 6–10–98

Accepted 12–4–99

431

98/012 AMA: First Proof 18 April, 19100

mammography, and microcalcifications were not reportedin any of the 22 patients. Similarly, a report byChandawarkar and Shinde15 indicated that ultrasound is assensitive as mammography in detecting palpable breastlesions, but they stressed that it was impossible to detectmicrocalcifications and certain intraductal cancers by ultra-sound.

The failure to detect early non-palpable breast lesionswith microcalcifications may be attributed to the relativeinsensitivity of ultrasound and to examiners’ unawarenessof the value of ultrasound in the identification of microcal-cifications. Recently, Yang and co-workers16 reported aseries of 82 invasive carcinomas, three ductal carcinomas insitu and four mucinous carcinomas. The sensitivity of high-resolution ultrasound in the identification of microcalcifi-cations within a mass in this study was 95%, with aspecificity of 87.8% and an accuracy of 91%. They con-cluded that the presence of microcalcifications is a reliablesign for diagnosing palpable malignancy and that it isdifficult to detect microcalcifications within ductal carcino-mas in situ, probably owing to the absence of a palpablemass. With the use of high-resolution ultrasound and ahigh-frequency broadband or confocal probe, however, wewere able to diagnose non-palpable breast cancers in youngwomen for whom mammography is usually not recom-mended17–19. We present our findings and the correlationbetween ultrasonography, mammography and histopatho-logy in 15 patients with non-palpable breast lesions.

MATERIALS AND METHODS

Fifteen patients (Table 1), aged between 26 and 59 years,with non-palpable breast lesions, were examined by both

breast ultrasonography and mammography. Fourteenpatients were recruited because microcalcifications weredetected or suspected by breast ultrasonography. Onepatient (patient 15) was recruited because mammographyshowed diffuse microcalcifications whereas ultrasono-graphy did not clearly demonstrate microcalcifications.None of these patients was at high risk for breast cancer.Risk factors included a previous personal history of breastcancer, a very strong family history of breast cancer, or aprevious biopsy showing lobular carcinoma in situ oratypical hyperplasia. When both ultrasound and mammo-graphy were requested, they were performed independentlyby different physicians at different times. For patientsyounger than 35 years, ultrasound was usually requestedfirst. Patient 5 is the only patient whose mammographicfindings were known before the ultrasound examination.

Breast ultrasound examination was performed via directskin contact over both breasts by use of a DiasonicsGateway series with a 10-MHz MI linear-array probe(Diasonics, Santa Clara, CA, USA) or an ATL HDI 3000scanner with an L10–5 (10–5 MHz) linear array probe(ATL, Bothell, WA, USA). Patients were in the supine posi-tion with the examined side elevated by a small pillow andthe ipsilateral arm raised above the head. The ultrasoundfindings were evaluated independently by two physicians(C.S.H. and J.H.L.) who were not at the time aware of themammographic findings of these patients with the excep-tion of patient 5. The presence of microcalcifications wasclassified into three groups: ‘positive’ if both examinersagreed, ‘suspicious’ and ‘negative’.

Mammography was performed with a SenographeDMR unit (General Electric Medical Systems, Milwaukee,WI, USA). Standard mediolateral oblique and craniocaudal

Ultrasonography Mammography – microcalcification

PatientAge

(years) Mass sizeMicro-

calcificationHypoechoic

patch or mass Morphology Number Pattern/size Pathology

1

234

5678

91011

12131415

30

282648

38485043

593955

43534846

—

——

1.4 × 1.2 cm

1.3 × 0.6 cm1.3 × 1.1 cm1.0 × 0.6 cm

—

———

cysts—

dilated duct—

yes

yesyesyes

yesyesyesyes

yessuspicioussuspicious

suspicioussuspicious

yesabsent

yes

yesyesyes

yesyesnono

yesnono

nonoyesno

g, l, b

gg, l, b

g

gggr

rg, lg

agg

numerous

numerousnumerousnumerous

numerous68

13

2numerousnumerous

numerous7

absentnumerous

diffuse (extensive)

clustered (1.8 cm)clustered (1.4 cm)clustered (1.5 cm)

clustered (0.8 cm)clustered (0.5 cm)clustered (0.2 cm)scattered

scattereddiffuse (extensive)diffuse (extensive)

clustered (0.2 cm)clustered (0.3 cm)nildiffuse (extensive)

IDC with extensive comedotype

DCIS, node negativeDCIS, comedo typeDCIS, comedo typeIDC (EIC > 25%), node

negativeDCIS, low gradeDCIS, intermediate gradeunknown (no biopsy)fibrocystic change, micro-

glandular adenosisunknown (no biopsy)DCIS, low gradefibrocystic change, florid

epithelial hyperplasiafibrocystic changefibrocystic changeunknown (no biopsy)fibrocystic change, ductal

hyperplasia, focal adenosis

g, granular; l, linear; b, branching; r, round; a, amorphous; numerous, number of microcalcifications ≥ 20; DCIS, ductal carcinoma in situ;IDC, infiltrating ductal carcinoma; EIC, extensive intraductal component

Table 1 Ultrasonographic, mammographic and pathological findings of 15 patients

98/012

Microcalcifications in breast ultrasonography Huang et al.

AMA: First Proof 18 April, 19100

432 Ultrasound in Obstetrics and Gynecology

a b

Figure 1 (a) Numerous microcalcifications demonstrated by ultrasonography in a palpable tumor as very bright dots with echogenicityhigher than that of the surrounding breast parenchyma. The microcalcification on ultrasound usually measures less than 1 mm and is notaccompanied by acoustic shadowing. (b) Ultrasonography revealing microcalcifications in tubular hypoechoic patches (arrows) (patient 1)

a

c

b

Figure 2 (a) Ultrasonography showing multiple microcalcifica-tions against ill-defined hypoechoic background (arrows), whichlooked like twinkling stars in a dark sky (patient 2). (b) Radio-graphy of the specimen (excised under ultrasound guidance) con-firmed the clustered microcalcifications (arrow) in the center.(c) Histopathology revealing high-grade comedo-type ductal carci-nomas in situ in 1 cm wide area of loosely packed ducts withincreased cellularity and necrosis. Hematoxylin and eosin, × 40

views were obtained. A magnified view was added whenclustered microcalcifications were detected.

Biopsy of breast lesions was carried out effectively viamammographically guided wire localization or ultrasound-guided localization without wire. By ultrasound, the lesioncould be localized directly with the patient lying on theoperating table with the breasts distributed evenly onthe chest wall. A wire was not necessary because a mark onthe overlying skin could help to locate the lesion. Biopsy ofthe lesion was then performed with an adequate resectionmargin. Tissues were processed and subjected to routinepathological examination.

RESULTS

High-resolution ultrasound showed microcalcifications tobe bright shining dots of greater echogenicity than that ofthe surrounding breast parenchyma, usually withoutacoustic shadows and measuring less than 1 mm in dia-meter. Since microcalcifications could be identified indifferent planes, some microcalcifications (bright dots) dis-appeared and others appeared when the probe was movedslowly across the skin. This made them appear like twin-kling stars in real-time ultrasonography (Figures 1 and 2).When microcalcifications were contrasted against hypo-echoic patches, they caught the examiner’s attentionquickly, making them hard to miss. Microcalcificationscould also be detected in small non-palpable masses(Figure 3).

The findings of ultrasonography, mammography andhistopathology for the 15 patients are shown in Table 1.Ten of the 15 patients (patients 1–9 and 14) were diag-nosed as having microcalcifications and four patients(patients 10–13) were suspected of having microcalcifica-tions by ultrasonography. The presence of microcalcifica-tions in 13 of the 14 patients (except patient 14) wasconfirmed by mammography. There was a discrepancy inthe number or degree of microcalcifications by ultrasono-graphy and mammography in patient 9 (Figure 4).

98/012

Microcalcifications in breast ultrasonography Huang et al.

AMA: First Proof 18 April, 19100

Ultrasound in Obstetrics and Gynecology 433

Among the nine patients (patients 1–9) positive on bothultrasound and mammography, patient 1 had numerousmicrocalcifications diffusely distributed in one quadrant ofthe right breast, and four (patients 2–5) were shown bymammography to have numerous microcalcifications clus-tered in an area smaller than 2 cm but larger than 0.5 cmin diameter. All five (patients 1–5) of these patients hadbiopsy-proven ductal carcinoma in situ with or withoutsmall infiltrating ductal carcinoma components (Figures 2and 3).

Two patients (patients 6 and 7) had only small numbers(< 10) of microcalcifications clustered in an area lessthan 0.5 cm in diameter by mammography. One of these(patient 6) revealed an intermediate ductal carcinoma insitu, and the other did not undergo surgery. The other twopatients among the nine (patients 8 and 9) revealed smallnumbers of microcalcifications diffusely scattered in onequadrant by mammography; one of them had a biopsy thatdid not reveal malignancy.

Of the four patients (patients 10–13) with suspiciousreadings on ultrasound, two had numerous, diffusely scat-tered granular or amorphous microcalcifications (patients10 and 11), and two had clustered microcalcifications(≤ 0.3 cm; patients 12 and 13), on mammography. Onlyone patient (case 10) had a low-grade ductal carcinoma insitu, and the remaining three had fibrocystic disease, asdetermined histologically.

Two patients (patients 14 and 15) showed discrepanciesbetween ultrasonographic and mammographic findings.Patient 14 was found by ultrasound to have microcalcifica-tions in a dilated duct, probably caused by floating debrisor crystals. Patient 15 had microcalcifications detected bymammography. Failure to show microcalcifications inpatient 15 was attributed to a relatively hyperechoic back-ground on ultrasound. A biopsy was not performed on thepatient (case 14) with a negative mammogram. No malig-nancy was reported in patient 15.

DISCUSSION

In this study, we demonstrated the usefulness of ultrasono-graphy and mammography in the detection of early breastcancer. Microcalcifications are one of the primary findings,if not the only finding, leading to suspicion of a cancer, andprovide the prime site for performing a mammographicallyguided biopsy1–6. Mammography still surpasses ultrasoundin sensitivity for the demonstration of microcalcifica-tions in non-palpable breast lesions. With the use of high-resolution ultrasound and a high-frequency broadband orconfocal probe, however, we demonstrated that ultrasoundwas able to detect microcalcifications in non-palpablebreast lesions when the microcalcifications were contrastedagainst a dark, hypoechoic background. In our previousexperience, microcalcifications were not demonstratedwell with a low-resolution scanner with a single, lower-frequency probe (7.5 MHz).

In this study, nine of 15 patients showed numerousclustered or diffuse microcalcifications on mammography.

a c

b

Figure 3 (a) Radiography of the specimen showing clustered microcalcifications (arrow) (patient 5). (b) Ultrasonography checked aftermammographically guided wire localization showed multiple microcalcifications in a hypoechoic mass (marked by X and +). The wire waspointed out by the mark (o) just below the wire tip. (c) Histopathology revealed a low-grade ductal carcinoma in situ with expanded ductscompacted in an area 0.8 cm wide. Hematoxylin and eosin, × 40

Figure 4 Ultrasonography showed suspected numerous micro-calcifications (arrow) but mammography showed only two scat-tered microcalcifications (patient 9). The discrepancy could beexplained by the difference in acoustic impedance between fattissue and suspensory ligaments or ductal tissue

98/012

Microcalcifications in breast ultrasonography Huang et al.

AMA: First Proof 18 April, 19100

434 Ultrasound in Obstetrics and Gynecology

Prominent microcalcifications with or without a mass wereconfirmed in five of the nine patients by ultrasound, and allfive of these patients had biopsy-confirmed malignancies.The microcalcifications could not be diagnosed confidentlyby ultrasound in the other four; three had diffuse microcal-cifications and one showed small clustered microcalcifica-tions. Among these four patients, only one had a low-grademalignancy and three had benign fibrocystic disease. Thus,microcalcifications confirmed by both ultrasonographyand mammography appeared to be sensitive and specific inthe detection of malignancy. Yang and colleagues16 drewthe same conclusion in palpable breast cancer. For non-palpable breast lesions, malignancy should be suspectedwhen microcalcifications are detected within a mass, orappear on ultrasound in an ill-defined hypoechoic back-ground, suggesting intraductal tumor growth.

Ultrasound is not free of false-positive or false-negativeresults. In this study, we encountered one patient (case 14)with false-positive ultrasound findings, which may havebeen due to the presence of floating debris or crystals thatmimic microcalcifications in the dilated duct. This patienthad suffered from mastitis, and symptoms and signs hadsubsided before the ultrasound examination. One patient(case 9) showed a discrepancy in the degree of microcalcifi-cations between mammography and ultrasound which mayhave been caused by ultrasound identifying fibrotic paren-chyma or suspensor ligaments as microcalcifications be-cause of their different acoustic impedance (Figure 4). Onepatient (case 15) gave a false-negative result on ultrasound.Prominent hyperechogenicity due to increased fibrosis anda lack of hypoechoic areas due to intraductal tumorgrowth, as in this patient, might impair the detection ofmicrocalcifications. Interestingly, the specimen radiographafter surgical biopsy showed relatively scattered microcalci-fications instead of the numerous microcalcifications asshown by mammography. The preoperative mammogrammight have compressed the scattered microcalcificationsinto one plane, thus making them appear numerous in eachview and centralized within a certain area. Therefore, fail-ure to detect microcalcifications by ultrasound could implythe absence of intraductal tumor growth or a scattereddistribution of microcalcifications.

In summary, our study indicates a positive correlationbetween malignancy and the presence of microcalcifica-tions detected by both mammography and ultrasound.Mammography, however, is not recommended in youngwomen under the age of 35 years because of their sensiti-vity to radiation exposure20,21. Also, because the false-negative rate of mammography is higher in young womenwith denser breast tissue than in older women17–19, ultra-sound may become preferable for detection of non-palpable lesions with or without microcalcifications inyoung patients. Ultrasound is also superior for the detec-tion of tumors in the breasts of pregnant or lactatingwomen, in surgically altered breasts, near prostheses, or atthe periphery of the breast. We strongly recommend the useof ultrasound for screening non-palpable breast cancers inwomen under 35 years, especially those with a familialhistory or who are at high risk of breast cancer. Nonethe-

less, we believe that microcalcifications detected by ultra-sound should be confirmed by mammography, as the lattertechnique can evaluate microcalcifications in greater detailin terms of morphology, number and extent.

ACKNOWLEDGEMENT

We thank Mei-Fan Chang, BS, for technical assistance.

REFERENCES1. Sickles EA. Mammographic features of 300 consecutive non-

palpable breast cancers. Am J Roentgenol 1986;146:661–32. Ciatto S, Cataliotti L, Distante V. Nonpalpable lesions

detected with mammography: review of 512 consecutive cases.Radiology 1987;165:99–102

3. de Lafontan B, Daures JP, Salicru B, Eynius F, Mihura J,Rouanet P, Lamarque JL, Naja A, Pujol H. Isolated clusteredmicrocalcifications: diagnostic value of mammography – seriesof 400 cases with surgical verification. Radiology 1994;190:479–83

4. Rosenberg AL, Schwartz GF, Feig SA, Patchefsky AS. Clini-cally occult breast lesions: localization and significance.Radiology 1987;162:167–70

5. Sickles EA. Breast calcifications: mammographic evaluation.Radiology 1986;160:289–93

6. Meyer JE, Kopans DB, Stomper PC, Lindfors KK. Occultbreast abnormalities: percutaneous preoperative needle local-ization. Radiology 1984;150:335–7

7. Sickles EA. Sonographic detectability of breast calcifications.Proc SPIE 1983;419:51–2

8. Yang WT, King W, Tang A, Metreweli C. Role of high fre-quency ultrasonography in the evaluation of palpable breastmasses in Chinese women: alternative to mammography.J Ultrasound Med 1996;15:637–44

9. Fornage BD, Coan JD, David CL. Ultrasound-guided needlebiopsy of the breast and other interventional procedures.Radiol Clin North Am 1992;30:167–85

10. Fornage BD, Sneige N, Faroux MJ, Andry E. Sonographicappearance and ultrasound-guided fine-needle aspirationbiopsy of breast carcinomas smaller than 1 cm. J UltrasoundMed 1990;9:559–68

11. Sneige N, Fornage BD, Saleh G. Ultrasound-guided fine-needleaspiration of nonpalpable breast lesions: cytologic and histo-logic findings. Am J Clin Pathol 1994;102:98–101

12. Parker SH, Jobe WE, Dennis MA, Stavros AT, Johnson KK,Yakes WF, Truell JE, Price JG, Kortz AB, Clark DG. Ultra-sonography-guided automated large-core breast biopsy.Radiology 1993;187:507–11

13. Gordon PB, Goldenberg SL, Chan NH. Solid breast lesions:diagnosis with ultrasonography-guided fine-needle aspirationbiopsy. Radiology 1993;189:573–80

14. Okamomto H, Ogawara T, Arihara F, Kobayashi K, Inoue S,Nagahori K, Yamamoto M, Sekikawa T, Matsumoto Y. Use-fulness of ultrasonography combined with conventional physi-cal examination in mass screening for breast cancer: aretrospective study of Yamanashi Health Care Center resultsfrom 1989 to 1994. Jpn J Cancer Res 1996;87:317–23

15. Chandawarkar RY, Shinde SR. Preoperative diagnosis ofcarcinoma of the breast: is a ‘cost-cutter’ algorithm tenable?J Surg Oncol 1997;64:153–8

16. Yang WT, Suen M, Ahuja A, Metreweli C. In vivo demonstra-tion of microcalcifications in breast cancer using high resolu-tion ultrasound. Br J Radiol 1997;70:685–90

17. Ciatto S, Bravetti P, Bonardi R, Rosselli del Turco M. The roleof mammography in women under 30. Radiol Med 1990;80:676–8

98/012

Microcalcifications in breast ultrasonography Huang et al.

AMA: First Proof 18 April, 19100

Ultrasound in Obstetrics and Gynecology 435

18. Harris VJ, Jackson VP. Indications for breast imaging inwomen under age 35 years. Radiology 1989;172:445–8

19. Evans WP. Breast masses: appropriate evaluation. Radiol ClinNorth Am 1995;33:1085–108

20. Edeiken S. Mammography and palpable cancer of the breast.Cancer 1988;61:263–5

21. Mettler FA, Upton AC, Kelsey CA, Ashby RN, RosenbergRD, Linver MN. Benefits versus risks from mammography: acritical reassessment. Cancer 1996;77:903–9

98/012

Microcalcifications in breast ultrasonography Huang et al.

AMA: First Proof 18 April, 19100

436 Ultrasound in Obstetrics and Gynecology