a Medical Imaging Centre, 6, place d’Italie, 75013 Paris, Franceb Gustave Roussy Institute, 114, rue Édouard-Vaillant, 94805 Villejuif, Francec Pathology Consulting Rooms, 19, rue de Passy, 75016 Paris, France
KEYWORDSBreast;Distortions;
Abstract Architectural distortions consist of convergence areas and local retractions at theborder of the gland. The authors examine the semiologic features of the distortions and their dif-ferent causes, together with their pathological anatomy correlations. The predominant benign
rchitectural distortions are due to defective connectiveissue harmony and include convergence areas and localetractions.
In the majority of cases they reflect a benign lesion,lthough after masses and microcalcifications, they are thehird leading appearance of cancers and are difficult toetect and manage. This article reviews the different causesf distortions, together with their anatomical-radiologicalorrelations.
onvergence areas consist of convergent spicules but witho central mass. They produce a star shaped appearance,ccasionally called a ‘‘black star’’ as they have no denseentre unlike the classical stellar appearance with a denseentre or mass with spiculated borders, known as a ‘‘whitetar’’ (Fig. 1).
A white star reflects centrifugal development of lesions,hich begin at the centre and almost invariably representn infiltrating carcinoma. The spicules are the centrifugalxtension of the lesions (Fig. 2).
The white star is therefore classified as BIRADS category
.
Conversely, in the majority of cases, a black star rep-esents centripetal development of lesions from retractionenerated by the centre of the lesion (Fig. 3). The black star
. Published by Elsevier Masson SAS. All rights reserved.
Figure 1. Mass with spiculated outlines (a) formed from a densecentre and spiculated outlines with very high likelihood of malig-nancy, to be classified as ACR category 5 and to be compared to theconvergence area (b) where spicules are present with no centralmass and which is classified as category 4.
Figure 4. Appearances suggestive of a left superior convergencearea on the oblique view (a) with no obvious centre: this is bettersp
t(
B
T
MTcl
Acma61rfi
Figure 2. White star (infiltrating ductal carcinoma): this is a cen-trifugal lesion invading the benign breast structures (b) peripherally.
is classified as category 4 as it is only malignant in 10 to 40%of cases [1].
The distinction between a black star and a white star isnot always straightforward on mammography if the densecentre is small (Fig. 4).
Retractions at the edge of the gland
These are difficult to diagnose and are identified by com-parison, examining for loss of continuity at the boundary of
Figure 3. Black star (radial scar): the benign breast structures (b)converge towards the centre containing fibro-elastosis.
DFsb
parobl
na
ptb
a
een on the postero-anterior view (arrow) (b) on which a mass isresent.
he gland. They may involve the superficial (Fig. 5) or deepFig. 6) boundary of the gland.
enign causes
he proliferative Aschoff body or radial scar
ammographic appearanceshese are generally found on routine mammography as aonvergence area without a dense centre, confirmed on aocal compression view (Fig. 7).
Some mammography appearances are suggestive of theschoff body: thin long spicules which are occasionallyurved or paired with radiotransparent bands, no palpableass even in superficial lesions and no dense centre (Fig. 8),
lthough these signs are not sufficiently specific as 30% to0% of cases are malignant [2]. Only 66.2% of a series of42 distortions suggestive of radial scarring [3] were in factadial scars, whereas 28.6% were malignant and 7% werebrocystic disease.
iagnosis and managementaced with these mammographic appearances, an ultra-ound is required. If this is positive, the Aschoff body cane diagnosed from a needle biopsy.
Management is controversial if pure radial scarring isresent: the conventional approach is to excise the lesions in 1/3 of cases, examination of the surgical specimeneveals either atypical lesions (15 to 20%) or an infiltratingr in situ carcinoma (15%) [3—5], which may not be seen oniopsy as it is occasionally located at the periphery of theesion.
Others suggest no treatment for an Aschoff body diag-osed on needle biopsy, if a sufficient number of samplesre taken and if no atypia is present [6,7].
Surgery, however, continues to be recommended for aroliferative Aschoff body because of the difficulty in moni-oring distortions and the risk of under-estimation on needle
iopsy [5].
If the ultrasound is normal, the question of diagnosisrises. Two approaches may be used:
136 B. Boyer, E. Russ
Figure 5. Local retraction of the superficial edge of the gland clearly seen on the postero-anterior view following the edge of the gland(arrow) (a) more difficult to see on the oblique view (arrow) (b): infiltrating lobular carcinoma.
Figure 6. Retraction of the deep edge of the gland with a convergence area visible on the postero-anterior (arrow) (a) and oblique (arrow)(b) views confirmed by local compression (arrow) (c): infiltrating ductal carcinoma.
Figure 7. Convergence area on the postero-anterior (arrow) (a) and oblique (arrow) (b) views confirmed by a local compression view (c).UT
ltrasound guided biopsy shows a proliferative Aschoff body or radial sche myoepithelial cells are revealed by immunohistochemistry using ant
ar (d), groping together tubules lined by a double basal cell layer.i-p63 antibodies (e).
Figure 10. Micro-biopsy (sclerosing adenosis): hyperplasia of theduct-lobule units, the organization of which is changed by fibrosisof the intra-lobular tissue. No atypia present.
Figure 11. Macro-biopsy (sclerosing adenosis): confirmation ofthe lesions described in Fig. 10. No atypia or neoplasia in the lobularroot.
Figure 8. Convergence area with no dense centre and with longcurved spicules.
• surgical biopsy: stereotactic biopsy raises the technicalproblem of targeting the abnormality because of its vari-ability between different views;
• MR to examine a target that can be biopsied prior tosurgery under MR guidance. If the MR is negative, as itis in 1/3 of cases of Aschoff bodies [8], monitoring maybe offered instead of surgery because of the high negativepredictive value of MR (97.4%) [9].
Sclerosing adenosis
Mammographic appearancesThis is characterized either by a mass or by foci of microcal-cifications [10,11] and more rarely by a convergence area(Fig. 9).
Management of sclerosing adenosisSclerosing adenosis is a benign, proliferative lesion, whichis often seen in the peri-menopausal period. It is not a high-risk lesion [10] although its histological diagnosis is difficultwith a risk of under-diagnosis. In addition, when histologyis carried out on a micro-biopsy (Fig. 10), this must be con-
firmed on a macro-biopsy (Fig. 11) [12]. If the diagnosis isconfirmed on a macro-biopsy and if no atypia is present,the patient may be monitored. If not, the general rule is toexcise the lesion.
Figure 9. Left supero-internal convergence area visible on the posterultrasound (c).
o-anterior (arrow) (a) and oblique (arrow) (b) views and found on
1
O
Tafi
M
I
MAfm
a[
DImos(
Fac
Fc
38
ther benign causes
hese are far more rare and include the Abrikossoff tumor, ubiquitously located granular tumor and the hyalinizedbroma.
alignant causes
nfiltrating lobular carcinoma
ammographic appearancesrchitectural distortion is a classical presenting appearanceor infiltrating lobular carcinoma, which is the most commonalignant cause of convergence areas and presents with this
tor(
igure 12. Sixty-nine year old patient. Screening mammography shownterior view (arrow) (a) which is seen on the oblique view (arrow) (arcinoma (d): fine rows of small non-cohesive tumor cells in a very poo
igure 13. Right superior convergence area visible on an oblique vionfirms distortion (b). Ultrasound is normal. MR (c) shows a mass with s
B. Boyer, E. Russ
ppearance in 16% to 20% of cases depending on the author13,14] (Fig. 12).
iagnosisf distortion is visible on both views, the diagnosis isade either by ultrasound guided needle micro-biopsy
r by MR vacuum-assisted macro-biopsy if the ultra-ound is negative. Surgical biopsy is occasionally requiredFig. 12).
Occasionally, distortion is only visible on one view, when
he diagnosis is more difficult. If the distortion is confirmedn a local compression view and ultrasound is negative, MR isequired to confirm actual distortion and provide a diagnosisFig. 13).
s a convergence area with no dense centre visible on the postero-b) and on the local view (c). Surgical biopsy: infiltrating lobularrly vascularized fibrous stroma.
ew (arrow) but not on the postero-anterior view. The local viewpiculated outlines (infiltrating lobular carcinoma).
Whilst intra-ductal carcinomas usually present with foci ofmicrocalcifications, they can also appear as a distortion(Fig. 14). They also represented 4% of asymptomatic distor-tions and almost 17% of palpable architectural distortions inPatterson’s series [15] (Table 1).
The calcifications seen in intra-ductal carcinomas maybe secretory (in which case the morphology is often roundor punctiform) or necrotic, which are then typically worm-like in appearance because of tumor proliferation within theducts onto which the calcifications form. Occasionally, thecentral necrosis does not calcify (Fig. 15), explaining theunusual mammographic appearance.
Infiltrating ductal carcinoma
This can also be seen as architectural distortion and raisesthe same diagnostic difficulties if no ultrasound appearancesare present (Fig. 6).
Table 1 Causes of architectural distortions depending on whe
Figure 14. Fifty-eight year old patient, screening mammography. Thgland (arrow) (a) confirmed by the local compression view (b) but not seof this diagnostic difficulty, MR was performed and showed segmental co5. Targeted ultrasound did not show any abnormality which could be bhigh-grade intra-ductal carcinoma.
139
onclusion
rchitectural distortions raise diagnostic and manage-ent difficulties as they do not always have ultrasound
ppearances and can generally not be biopsied underammography. In view of their high predictive value foralignancy (30%), practitioners should be prepared to pro-
eed as far as MR or even surgical biopsy to obtain aiagnosis. The predominant benign lesions are the prolif-rative Aschoff body and the main malignant causes arenfiltrating lobular carcinomas.
TAKE-HOME MESSAGES
• Architectural distortions involve convergence areas
ther or not they are palpable [15].
ia (%)Intra-ductalcarcinoma (%)
Infiltratingcarcinoma (%)
4.7 2.316.9 15.7
e mammogram shows local retraction of the superior edge of theen on the postero-anterior view (c). Ultrasound is normal. Becausentrast uptake in the suspected area (d), classified as ACR categoryiopsied. The biopsy was then performed under MR and showed a
140 B. Boyer, E. Russ
F ratioc Math
D
Tc
R
[
[
[
[
[
[15] Patterson JA, Scott M, Anderson N, Kirk SJ. Radial scar, com-
igure 15. Biopsy (in situ ductal carcinoma): intra-ductal prolifealcification and of high nuclear grade (a and b) (with thanks to MC
• They are classified as BIRADS, category 4.• The most common cause is the proliferative Aschoff
body.• If an Aschoff body is diagnosed, surgical excision is
still the general rule because of the possibility of aco-existent carcinoma.
• Sclerosing adenosis can just be monitored if it hasbeen diagnosed by a macro-biopsy and does notcontain atypia.
• The most common malignant cause of architecturaldistortions is infiltrating lobular carcinoma.
• Intra-ductal carcinomas may produce appearances ofpure distortions with no calcification in the absenceof any calcified necrosis.
isclosure of interest
he authors declare that they have no conflicts of interestoncerning this article.
eferences
[1] Digabel-Chabay C, Labbe-Devilliers AC, Meingan C, RicaudCouprie CM. Distorsions architecturales et difficultés diagnos-tiques, 14. Le Sein; 2004. p. 100—8.
[2] Tardivon A. Les images détectées. In: France SV, editor. Ledépistage du cancer du sein un enjeu de santé publique. Paris:Seradour B.; 2004. p. 89—104.
[3] Farshid G, Rush G. Assessment of 142 stellate lesions withimaging features suggestive of radial scar discovered duringpopulation-based screening for breast cancer. Am J Surg Pathol
2004;28(12):1626—31.
[4] Morgan C, Shah ZA, Hamilton R, Wang J, Spigel J, Deleon W,et al. The radial scar of the breast diagnosed at core needlebiopsy. Proc (Bayl Univ Med Cent) 2012;25(1):3—5.
n of cohesive tumor cells with comedome-like architecture and noieu).
[5] Osborn G, Wilton F, Stevens G, Vaughan-Williams E, Gower-Thomas K. A review of needle core biopsy diagnosed radialscars in the Welsh Breast Screening Programme. Ann R Coll SurgEngl 2011;93(2):123—6.
[6] Brenner RJ, Jackman RJ, Parker SH, Evans WP, 3rd, PhilpottsL, et al. Percutaneous core needle biopsy of radial scars ofthe breast: when is excision necessary? AJR Am J Roentgenol2002;179(5):1179—84.
[7] Sohn VY, Causey MW, Steele SR, Keylock JB, Brown TA. Thetreatment of radial scars in the modern era–surgical excision isnot required. Am Surg 2010;76(5):522—5.
[8] Linda A, Zuiani C, Londero V, Cedolini C, Girometti R, Baz-zocchi M. Magnetic resonance imaging of radial sclerosinglesions (radial scars) of the breast. Eur J Radiol 2012;81(11):3201—7.
[9] Linda A, Zuiani C, Furlan A, Lorenzon M, Londero V, GiromettiR, et al. Nonsurgical management of high-risk lesions diagnosedat core needle biopsy: can malignancy be ruled out safely withbreast MRI? AJR Am J Roentgenol 2012;198(2):272—80.
10] Taskin F, Köseoglu K, Unsal A, Erkus M, Ozbas S, KaramanC. Sclerosing adenosis of the breast: radiologic appearanceand efficiency of core needle biopsy. Diagn Interv Radiol2011;17(4):311—6.
11] Gill HK, Ioffe OB, Berg WA. When is a diagnosis ofsclerosing adenosis acceptable at core biopsy? Radiology2003;228(1):50—7.
12] Li JL, Wang ZL, Su L, Liu XJ, Tang J. Breast lesions with ultra-sound imaging-histologic discordance at 16-gauge core needlebiopsy: can re-biopsy with 10-gauge vacuum-assisted systemget definitive diagnosis? Breast 2010;19(6):446—9.
13] Le Gal M, Ollivier L, Asselain B, Meunier M, Laurent M, Vielh P,et al. Mammographic features of 455 invasive lobular carcino-mas. Radiology 1992;185(3):705—8.
14] Evans WP, Warren Burhenne LJ, Laurie L, O’ShaughnessyKF, Castellino RA. Invasive lobular carcinoma of the breast:mammographic characteristics and computer-aided detection.Radiology 2002;225(1):182—9.
plex sclerosing lesion and risk of breast cancer. Analysis of175 cases in Northern Ireland. Eur J Surg Oncol 2004;30(10):1065—8.
