Wo m e n ’s I m a g i n g • C l i n i c a l O b s e r v a t i o n s Mesurolle et al. Sonography and Mammography of Breast Hemangioma
Downloaded from www.ajronline.org by 37.44.207.133 on 01/22/17 from IP address 37.44.207.133. Copyright ARRS. For personal use only; all rights reserved
Women’s Imaging Clinical Observations
W O M E N ’S IMAGING
Sonographic and Mammographic Appearances of Breast Hemangioma
Benoît Mesurolle1 Vitaly Sygal1 Lucie Lalonde2 André Lisbona3 Michel-Pierre Dufresne 4 Jean H. Gagnon1 Ellen Kao1
OBJECTIVE. The purpose of our study was to retrospectively evaluate the clinical, imag ing, and pathologic findings of breast hemangiomas in 16 patients. CONCLUSION. A mass displaying an oval or lobular shape with well-circumscribed or microlobulated margins on mammography and sonography, and in a superficial location, should alert the radiologist to the possible diagnosis of hemangioma. Imaging-guided biopsy appears sufficiently reliable to rule out any malignant or premalignant component and to avoid a surgical excision if doing so is clinically appropriate.
Mesurolle B, Sygal V, Lalonde L, et al.
emangiomas have been described in many anatomic locations, but there are few recorded examples of hemangiomas arising primar ily in the breast [1, 2]. The mammographic and sonographic appearances of breast he mangiomas have been described in a few case reports [3–7] but, to our knowledge, no case series including long-term follow-up imaging have been published. A search of the surgical pathology records at four institutions of breast biopsies per formed from 2000 to 2005 identified 16 pa tients with pathologically proven breast he mangiomas. Clinical, mammographic, and sonographic features were reviewed, with particular emphasis on sonographic findings.
Keywords: breast tumor, hemangioma, mammography, sonography DOI:10.2214/AJR.07.3153 Received September 14, 2007; accepted after revision January 29, 2008. 1 Department of Radiology, Cedar Breast Clinic, McGill University Health Center, Royal Victoria Hospital, 687 Pine Ave. W, Montréal, PQ H3G 1A4, Canada. Address correspondence to B. Mesurolle (
[email protected]). 2 Département de Radiologie, Centre d’Imagerie du Sein, Centre Hospitalier de l’Université de Montréal (CHUM), Montréal, PQ, Canada. 3
Department of Radiology, Sir Mortimer B. Davis-Jewish General Hospital, Montréal, PQ, Canada. 4
Département de Radiologie, Hôpital MaisonneuveRosemont, Montréal, PQ, Canada. CME This article is available for CME credit. See www.arrs.org for more information. WEB This is a Web exclusive article. AJR 2008; 191:W17–W22 0361–803X/08/1911–W17 © American Roentgen Ray Society
AJR:191, July 2008
H
Materials and Methods Study Population The combined data were obtained from four large institutions affiliated with two academic medical cen ters (McGill University and University of Montréal). A total of 23 pathologically proven hemangiomas diagnosed by either sonographically guided, stereotactically guided, or surgical excision biopsy performed between January 2000 and December 2005 were identified. Seven patients for whom no imaging was available were excluded. Sixteen wo men (age range, 43–75 years; mean age, 59.7 years) who met these criteria are the subject of this article. Two radiologists, specialists in breast imaging, retrospectively reviewed in consensus the mammo graphic and sonographic features of the hemangi omas. Approval by the institutional review board was not required for this retrospective analysis.
Mammographic Findings Mammography was performed in two standard imaging planes, mediolateral oblique and cranio caudal, using dedicated film-screen mammographic equipment. Mammographic characteristics were assessed according to the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) [8]. Mammographic findings are shown in Figures 1–4.
Sonographic Findings Breast sonography was performed to evaluate specific abnormalities discovered either at clinical examination or on mammography. All 16 patients underwent a breast sonographic examination (Figs. 1–4), but a mass was sonographically detected in only nine patients (Figs. 1–3). The sonographic char acteristics of the nine masses were assessed accord ing to the American College of Radiology BI-RADS criteria [8]. Studies were reviewed on hard-copy films (five patients) or on PACS images (four patients).
Core Needle Biopsy Fifteen patients underwent an imaging-guided core needle biopsy. One patient underwent surgical excision without preoperative biopsy. Biopsy of masses was performed with either sonographic or stereotactic guidance. If the mass was clearly identi fied sonographically, the biopsy was performed under sonographic guidance. If not, the biopsy was performed under stereotactic guidance. Stereotactic biopsies were performed with dedicated prone stereotactic biopsy tables (n = 7). Sonographically guided biopsy was performed with patients in the supine position and with high-resolution sonographic equipment (15L8W broadband transducer, Sequoia,
W17
Downloaded from www.ajronline.org by 37.44.207.133 on 01/22/17 from IP address 37.44.207.133. Copyright ARRS. For personal use only; all rights reserved
Mesurolle et al.
A
B
C
Fig. 1—Asymptomatic 75-year-old woman. A, Right mediolateral oblique mammogram shows 0.8-cm mass in superior aspect of breast. Mass (arrow) displays lobular shape and circumscribed margins. B, Close-up view of mass, also in right mediolateral oblique orientation. C, Breast sonogram shows 0.8-cm oval solid mass (arrows) with circumscribed margins, parallel orientation, abrupt interface, neutral sound transmission, and complex echotexture. Surgical excision revealed hemangioma, cavernous type.
A
B
C
Fig. 2—Asymptomatic 62-year-old woman. A and B, Left craniocaudal screening mammogram shows 1.2-cm lobular mass (arrow) with microlobulated margins in upper inner aspect of breast. Mass (arrow) shows interval increase in size compared with previous mammogram 8 years earlier (B). C, Breast sonogram using tissue harmonic imaging shows 1.1-cm subtle mass superficially situated against fatty background. Lesion is ovoid with microlobulated margins, hypoechoic echotexture, parallel orientation, and mild posterior acoustic shadowing (arrows). Sonographically guided core biopsy using 14-gauge needle yielded cavernous hemangioma. Follow-up imaging after 2 years (not shown) confirmed stability of lesion.
W18
AJR:191, July 2008
Downloaded from www.ajronline.org by 37.44.207.133 on 01/22/17 from IP address 37.44.207.133. Copyright ARRS. For personal use only; all rights reserved
Sonography and Mammography of Breast Hemangioma
A
B
C
Fig. 3—46-year-old woman with palpable mass in lateral aspect of left breast. A and B, Left craniocaudal close-up image (A) and left mediolateral oblique image (B) show mass (arrow) displaying irregular shape and microlobulated margin and having no microcalcifications. C, Breast sonogram shows 1.5-cm solid mass with irregular shape and indistinct margins (arrows), parallel orientation, neutral sound transmission, and hyperechoic echotexture. Sonographically guided biopsy using 14-gauge needle yielded capillary hemangioma. Surgical excision confirmed hemangioma, capillary type.
A
B
Fig. 4—Asymptomatic 58-year-old woman with screening-detected mass. Mammogram of right breast showed 1-cm mass. A, Radiograph obtained during stereotactic biopsy with 14-gauge needle in place shows round mass with circumscribed margins and punctuate microcalcifications. Breast sonogram did not identify any mass. Stereotactically guided biopsy using 14-gauge needle yielded cavernous hemangioma. B, Follow-up right craniocaudal close-up mammogram obtained 2 years later shows 0.7-cm mass (arrow).
Siemens Medical Systems; high-frequency Matrix transducer PLT1204AX, Aplio, Toshiba; model 700, GE Healthcare; and HDL, ATL) (n = 8). Generally, 10- to 15-MHz transducers were used. Fourteengauge automated core biopsy needles were used for stereotactically and sonographically guided biopsies.
Surgical Excision Of the 15 patients who underwent core needle biopsy, three underwent subsequent surgical excision. One patient underwent surgical excision without preoperative biopsy.
Pathologic Findings The prospective pathology reports by the institutional pathologists at the time of core needle
AJR:191, July 2008
biopsy (12 patients with no subsequent surgery) and at the time of the surgical excision (four patients, including three with previous core needle biopsy) were considered to indicate the definitive pathologic results.
Results Clinical Data Patients were referred to our institutions be cause of a palpable mass (n = 2), a screeningdetected mass (n = 8), or an increase in size of a previously screening-detected mass (n = 6). Mammographic Findings In the 16 patients studied, mammographic parenchymal density was BI-RADS class 3
(heterogeneously dense) in three patients (18.75%), class 2 (scattered fibroglandular densities) in five (31.25%), and class 1 (fatty) in eight (50%). One mammogram was normal (class 2), the mass being palpable and situated in the axillary region. Fifteen mammograms showed a mass (mean size, 1.2 cm; range, 0.6–2.5 cm) (Fig. 5). Mammographic findings are summarized in Table 1. The most com mon mammographic features were masses with a lobular shape (n = 8), circumscribed (n = 8) or microlobulated (n = 7) margins, and density equal to that of the breast parenchyma (n = 14) (Figs. 1 and 2). One patient had punc tate microcalcifications in a mass (Fig. 4). Sonographic Findings Sonographic findings are summarized in Table 2. Nine masses were identified at sono graphic examination (Figs. 1–3). The mean sonographic size was 1.3 cm (range, 0.6–3.2 cm) in the largest diameter (Fig. 6). The most common sonographic features of hemangiomas were oval shape (n = 8); circumscribed margins (n = 5); superficial with an abrupt interface (n = 9); and nonhyperechoic (i.e., hypoechoic, isoechoic, or complex) echotextures (n = 6). The two lesions displaying indistinct margins had a hyperechoic echotexture (Fig. 3). Pathologic Findings Final pathologic examination identified 14 cavernous-type hemangiomas and two capil lary-type hemangiomas. Among the four pa tients who underwent excisional biopsy, no malignancy was diagnosed. Surgical exci sion confirmed the diagnosis of hemangioma
W19
Mesurolle et al. Fig. 5—Scatterplot shows size of 15 masses that were visible on mammography.
3.0
Lesion Size (cm)
Downloaded from www.ajronline.org by 37.44.207.133 on 01/22/17 from IP address 37.44.207.133. Copyright ARRS. For personal use only; all rights reserved
2.5 2.0 1.5 1.0 0.5 0.0 0
1
2
3
4
5
6
7
9 10 11 12 13 14 15 16
8
No. of Lesions
TABLE 1: Mammographic Characteristics of 15 Hemangiomas According to BI-RADS Lexicon [8] Mass Characteristic
No.
%
Round
1
6.7
Oval
5
33.3
Lobular
8
53.3
Irregular
1
6.7
Circumscribed
8
53.3
Microlobulated
7
46.3
Obscured
0
0
Indistinct
0
Spiculated
0
Shape
TABLE 2: Sonographic Characteristics of Nine Hemangiomas According to BI-RADS Lexicon [8] Mass Characteristic
No.
%
Shape Irregular
1
11
Round
0
0
Oval
8
89
Circumscribed
5
56
Microlobulated
2
22
Angular
0
0
Spiculated
0
0
0
Indistinct
2
22
0
Orientation Parallel
9
100
Not parallel
0
0
Margins
Density
Margins
High
1
6.7
Equal
14
93.3
Lesion boundary
Low
0
0
Abrupt interface
9
100
Fat
0
0
Echogenic halo
0
0
Calcifications
1
6.7
Echo pattern Hypoechoic
2
22
in three patients and established the diagno sis in one patient. Surveillance Four patients (25%) underwent surgical ex cisional biopsy (Figs. 1 and 3). Eight patients (50%) underwent mammographic surveillance and were followed up for a mean of 3.7 years (range, 2–6 years) (Figs. 2 and 4). Among these eight patients, six breast masses were mammo graphically stable and two showed a decrease in size. Four of the 16 patients (25%) had no mammographic surveillance. Among a subgroup of patients referred for increase in size of a previously screening-
W20
Isoechoic
1
11
Hyperechoic
3
33
Complex
3
33
detected mass (n = 6), two had no mammo graphic surveillance, three showed mammo graphically stable masses (3–6 years), and one showed a decrease in size of the mass on mammography. Discussion Vascular tumors of the breast are rare le sions. Three main entities have been reported:
hemangiomas, angiolipomas, and angiosarco mas [9]. Hemangiomas are benign vascular tumors of two common types (capillary or cav ernous) that are based on the size of the vessels involved. Breast hemangiomas are found in 1.2% of mastectomy specimens and 11% of postmortem specimens of the female breast [9]. In our case series we focused our observa tions exclusively on the benign form of vascu lar tumors of the breast, the hemangiomas. Most of the breast hemangiomas in our se ries appeared on mammography as oval or lobular isodense masses with well-circum scribed margins, similar to the classic ap pearance of a hemangioma noted in previous case reports [3, 6]. These findings are not sufficiently specific and might explain the low incidence of diagnosed and recognized breast hemangiomas, a significant number probably being classified as BI-RADS cate gory 3 and not differentiated from fibroade nomas or cysts [10–12]. Although these lesions were superficial in location (subdermal or in the subcutaneous tissues), this location made it difficult to visu alize them on regular craniocaudal and me diolateral oblique views; tangential views are necessary to prove the superficial nature of the mass. Only one patient showed a mass dis playing benign round calcifications that were presumably related to phleboliths. A few cases of hemangiomas displaying suspicious mam mographic or clinical features have been re ported [5, 13]. One article reports a large pal pable mass measuring 12 cm, occupying the lower quadrant of the left breast, showing red dish discoloration of the overlying skin and nipple areolar complex, and mimicking an in flammatory carcinoma [5]. A second article reported a spiculated mass on mammography, leading the radiologists to classify the mass as BI-RADS category 5 [13]. In our series, no such features were noted. Only one patient had calcifications in the mass. Fine, punctu ate, or phlebolith-type calcifications have been reported in breast hemangiomas [3]. Although calcifications are classic, their fre quency has not been established. They can be of the venous type because of phleboliths, or they may be coarse [3]. As noted in several articles, the sonograph ic appearance of breast hemangioma is vari able. The shape and margins are usually con sistent with benign lesions and similar to the mammographic features. The lesions display an ovoid shape with well-circumscribed mar gins (Fig. 1). Microlobulations or indistinct margins are less often seen; the latter have
AJR:191, July 2008
Sonography and Mammography of Breast Hemangioma Fig. 6—Scatterplot shows size of nine masses that were visible on sonography.
agement of patients diagnosed with breast hemangioma by imaging-guided biopsy if no discord exists among mammographic, sono graphic, and pathologic findings. In conclusion, a mass displaying an oval or lobular shape, with well-circumscribed or mi crolobulated margins on mammography and sonography, and showing a superficial location should alert the radiologist to a possible diag nosis of hemangioma. Imaging-guided biopsy appears to be sufficiently reliable to rule out any malignant or premalignant component and to avoid surgical excision in the absence of radiologic–pathologic discordance.
were presumed to be in the subcutaneous tis sue even though the superficial pectoralis fascia was not visualized. Complete excision is usually recommended when a hemangioma is diagnosed at biopsy. The two main reasons for excision would be to exclude the possibility of an underlying angiosarcoma and to avoid progression to an giosarcoma. On one hand, imaging-guided biopsy appears sufficiently reliable to rule out any malignant or premalignant component; therefore, if no pathologic–radiologic discor dance is noted, the diagnosis of angiosarcoma can probably be excluded. On the other hand, progression of hemangioma to angiosarcoma has been reported but is controversial [16]. Because microscopic hemangioma shows a relatively high incidence at autopsy (11%) [17] and because the incidence of angiosar coma is extremely low, progression of he mangiomas to angiosarcoma is extremely rare if it exists at all. Hence, although needle biopsies may pro duce false-negative results, we consider that a superficial breast mass displaying the clas sic mammographic and sonographic appear ances of a hemangioma and diagnosed as a hemangioma (cavernous or capillary type, without atypia) by imaging-guided core bi opsy, can be followed up with periodic imag ing surveillance without surgical excision. Because of the established validity of mam mographic surveillance, follow-up with mammography alone at 12-month intervals for the first 2 years should be sufficient to show stability. Our study has some limitations. The small size of the population and the retrospective nature of the study probably limit our results. Nevertheless, although 25% of patients had no follow-up imaging, we emphasize the benefit of imaging surveillance in the man
1. Jozefczyk MA, Rosen PP. Vascular tumors of the breast. II. Perilobular hemangiomas and heman giomas. Am J Surg Pathol 1985; 9:491–503 2. Rosen PP. Vascular tumors of the breast. V. Non parenchymal hemangiomas of mammary subcu taneous tissues. Am J Surg Pathol 1985; 9:723– 729 3. Webb LA, Young JR. Case report: haemangioma of the breast—appearances on mammography and ultrasound. Clin Radiol 1996; 51:523–524 4. Chung SY, Oh KK. Mammographic and sono graphic findings of a breast subcutaneous heman gioma. J Ultrasound Med 2002; 21:585–588 5. Gopal SV, Nayak P, Dharanipragada K, Krishna machari S. Breast hemangioma simulating an in flammatory carcinoma. Breast J 2005; 11:498– 499 6. Siewert B, Jacobs T, Baum JK. Sonographic eval uation of subcutaneous hemangioma of the breast. AJR 2002; 178:1025–1027 7. Glazebrook KN, Morton MJ, Reynolds C. Vascular tumors of the breast: mammographic, sonographic, and MRI appearances. AJR 2005; 184:331–338 8. American College of Radiology. Breast imaging reporting and data system: breast imaging atlas, 4th ed. Reston, VA: American College of Radiol ogy, 2003 9. Lesueur GC, Brown RW, Bhathal PS. Incidence of perilobular hemangioma in the female breast. Arch Pathol Lab Med 1983; 107:308–310 10. Graf O, Helbich TH, Hopf G, Graf C, Sickles EA. Probably benign breast masses at US: is follow-up an acceptable alternative to biopsy? Radiology 2007; 244:87–93 11. Sickles EA. Probably benign lesions: when should follow-up be recommended and what is the optimal follow-up protocol? Radiology 1999; 213:11–14 12. Graf O, Helbich TH, Fuchsjaeger MH, et al. Followup of palpable circumscribed noncalcified solid breast masses at mammography and US: can biopsy be averted? Radiology 2004; 233:850– 856 13. Mariscal A, Casas JD, Balliu E, Castella E. Breast
3.5
Lesion Size (cm)
Downloaded from www.ajronline.org by 37.44.207.133 on 01/22/17 from IP address 37.44.207.133. Copyright ARRS. For personal use only; all rights reserved
3.0 2.5 2.0 1.5 1.0 0.5 0.0 0
1
2
3
4
5
6
7
8
9
10
No. of Lesions
References been associated with hyperechoic lesions. The echotexture of hemangiomas is variable; one third display a hyperechoic echotexture and two thirds display an isoechoic (to the fat), hy poechoic, or complex echotexture. Interest ingly, the echotexture is complex in a signifi cant number of cases (Fig. 1). This relative echotextural heterogeneity may be related to the presence of multiple small vascular chan nels seen pathologically in hemangiomas (large blood-filled spaces or sinuses being seen in cavernous hemangiomas) [3, 7]. The isoechoic or slightly hypoechoic appearances likely explain the difficulty in identifying such lesions against a fatty background. For this reason, hemangiomas are less conspicu ous sonographically than they are clinically or mammographically [7]. This characteris tic might account for the invisibility of the lesions on sonography in seven of our pa tients, despite these lesions being readily vis ible on mammography as well-defined mass es surrounded by fatty tissue. In such cases, the knowledge of the superficial location of the lesion alerts the radiologist to focus the examination on the subcutaneous region (ad equate focus positioning), which allows bet ter identification of the lesions. In addition, the use of specific settings on sonography, such as tissue harmonic imaging, increases the visibility of such lesions against a fatty background [14] (Fig. 2). Superficial location is another characteristic of these lesions. The subcutaneous location may theoretically help to select the correct diagnosis from the differential diagnosis, which includes hematoma, hemangioma, and sebaceous cyst [6]. To ascertain the subcuta neous location, the mass must be located an terior to the superficial pectoralis fascia [15], which is often difficult to visualize. In our study, the lesions appeared superficial and
AJR:191, July 2008
W21
Mesurolle et al.
Downloaded from www.ajronline.org by 37.44.207.133 on 01/22/17 from IP address 37.44.207.133. Copyright ARRS. For personal use only; all rights reserved
hemangioma mimicking carcinoma. Breast 2002; 11:357–358 14. Mesurolle B, Helou T, El-Khoury M, Edwardes M, Sutton EJ, Kao E. Tissue harmonic imaging, frequency compound imaging, and conventional imaging: use and benefit in breast sonography. J
Ultrasound Med 2007; 26:1041–1051 15. Fornage BD, McGavran MH, Duvic M, et al. Im aging of the skin with 20-MHz US. Radiology 1993; 189:69–76 16. Rupec M, Batzenschlager I. Pseudoangiosarcoma (Masson): a histological study [in German]. Z
Hautkr 1981; 56:1360–1363 17. Hoda SA, Cranor ML, Rosen PP. Hemangiomas of the breast with atypical histological features: fur ther analysis of histological subtypes confirming their benign character. Am J Surg Pathol 1992; 16:553–560
F O R YO U R I N F O R M AT I O N
This article is available for CME credit. See www.arrs.org for more information.
W22
AJR:191, July 2008
