DOI:10.2214/AJR.06.0099
AJR 2007; 188:1568-1572
© American Roentgen Ray Society
High-Resolution MRI in Detecting Subareolar Breast Abscess
Peifen Fu1,2,
Yasuyuki Kurihara1,
Yoshihide Kanemaki1,
Kyoko Okamoto1,
Yasuo Nakajima1,
Mamoru Fukuda3 and
Ichiro Maeda4
1 Department of Radiology, School of Medicine, St. Marianna University, 2-16-1
Sugao, Miyamae-ku, Kawasaki City, Kanagawa Prefecture 216-8511, Japan.
2 Department of Surgery, First Affiliated Hospital, School of Medicine, Zhejiang
University, Hangzhou, China.
3 Department of Surgery, School of Medicine, St. Marianna University, Kawasaki
City, Kanagawa Prefecture, Japan.
4 Department of Pathology, School of Medicine, St. Marianna University, Kawasaki
City, Kanagawa Prefecture, Japan.
Received January 18, 2006;
accepted after revision May 30, 2006.
Address correspondence to Y. Kurihara
(y4kuri{at}marianna-u.ac.jp).
Abstract
OBJECTIVE. Because subareolar breast abscess has a high recurrence
rate, a more effective imaging technique is needed to comprehensively
visualize the lesions and guide surgery. We performed a high-resolution MRI
technique using a microscopy coil to reveal the characteristics and extent of
subareolar breast abscess.
CONCLUSION. High-resolution MRI has potential diagnostic value in
subareolar breast abscess. This technique can be used to guide surgery with
the aim of reducing the recurrence rate.
Keywords: abscess • breast • MRI
Introduction
Although subareolar breast abscess was first described by Zuska et al.
[1] in 1951 as "fistulas
of lactiferous ducts," it remains a troublesome entity affecting
nonpuerperal women. It has a tendency to recur and form fistulas. Once
subareolar breast abscess becomes intractable, cosmetic results are
disappointing.
Subareolar breast abscess is located in the retroareolar area or within 1
cm of the areola. Clinical features vary depending on the stage. Initially,
mastalgia, subareolar lump, and overlying skin inflammation are noted with or
without nipple discharge. Later, a fluctuant abscess is palpated, which will
yield pus spontaneously or on incision. If the abscess becomes chronic,
fistula and nipple inversion frequently occur. Because of the close
relationship between subareolar breast abscess and the nipple–areola
complex, it is difficult to fully assess the lesion by conventional radiologic
examinations such as mammography and sonography. Insufficient preoperative
planning may be one of the reasons for inadequate surgical therapy and
subsequent recurrence. We therefore propose a high-resolution MRI technique
using a microscopy coil to reveal the characteristics and extent of subareolar
breast abscess and to better guide surgical therapy. To our knowledge, no
literature has been published on the application of this technique to the
detection of subareolar breast abscess. To evaluate the potential diagnostic
value of high-resolution MRI, we report the high-resolution MRI findings of 12
patients with subareolar breast abscess. Enhancement patterns were
investigated, and architectural features of these lesions on MRI were compared
with those seen on mammography and sonography.
Materials and Methods
This was a retrospective study of 12 patients who had recurrent history of
subareolar breast abscess that was under poor control. The patients underwent
surgery in our institution between September 2002 and April 2005. Mammography
was performed in five patients, and sonography and high-resolution MRI were
performed in all cases to confirm the diagnosis and define the extent of the
inflammatory focus. The medical records were reviewed to gather clinical and
radiologic data and surgical details. Follow-up information was collected from
case notes.
MRI was performed using a 1.5-T Intera Master unit (Philips Medical
Systems). All patients underwent imaging in the prone position. A microscopy
coil with a diameter of 4.7 cm was used as a surface coil to acquire signal
data from the nipple–areola complex
(Fig. 1). Construction of this
coil allowed local acquisition of the signal with a high signal-to-noise
ratio. This high signal-to-noise ratio was used to increase spatial resolution
by acquiring a 512 x 256 matrix for a 7-cm field of view, resulting in a
0.137-mm in-plane resolution, much higher than that of existing coils
(0.3–0.7 mm). An inherent property of these small coils, which display
high local sensitivity, is that signal yield at greater distances from the
coil is low. This problem was overcome by using a postprocessing technique
that provided a correction, resulting in uniform signal level. The technique
was performed by acquiring auxiliary signals from the small coil. This
technique was accomplished using CLEAR (constant level appearance) software
(Philips Medical Systems).
Axial T2-weighted (TR/TE, 1,500/120) turbo spin-echo and axial T1-weighted
(43/8.7; flip angle, 50°) 3D fast-field echo imaging sequences were
obtained using a fat-saturation technique and the CLEAR algorithm with a slice
thickness of 1.6 mm (0.8-mm overlap) and a 512 x 256 matrix. Scanning
time for both T1-weighted and T2-weighted images was approximately 5 minutes.
Gadodiamide hydrate (Omniscan, Daiichi Pharmaceutical) was administered IV as
a bolus injection at a dose of 0.1 mmol/kg body weight over 10 seconds. Serial
dynamic MRI was performed before injection of the contrast agent (unenhanced
imaging) and at 30 seconds (early phase imaging) and 5 minutes (delayed phase
imaging) after the start of the bolus injection.
All MR images were evaluated in consensus interpretations by two
radiologists experienced in breast imaging. The MRI data were collected to
facilitate preoperative planning by revealing abscess size and location and
characteristics such as nipple inversion and presence of abscess cavities,
fistulas, dilated lactiferous ducts, and inflammatory signs around the
abscess.
Results
The patients were all women ranging in age from 22 to 50 years, with a mean
age of 30.6 years. Diagnosis of subareolar breast abscess was confirmed by
histopathology.
Mammography was performed in five patients to exclude malignancy; findings
were normal in all but one, who exhibited a focal asymmetric density without a
distinct margin. Malignant calcification was not evident in any of the
patients. In 10 (83%) of the patients, sonography showed irregular lesions;
four of these lesions were well defined and six were poorly defined. All
lesions had poor or heterogeneous internal echoes. Two patients who had fine
fistulas along the nipple but no palpable masses exhibited no specific
sonographic findings.
High-resolution MRI detected all the inverted nipples, abscess cavities,
and fistulas in these patients. Of the 12 patients, nipple inversion was found
in nine (75%), abscess cavities in three (25%), fistulas in four (33%), and
both abscess cavities and fistulas in five (42%). In one patient, two
bilateral fistulas were also visualized (Fig.
2A,
2B,
2C). The location and extent of
the abscesses were well shown by MRI (Table
1). Findings of unenhanced T1-weighted images and T2-weighted
images were nonspecific, whereas contrast-enhanced dynamic T1-weighted scans
showed irregular heterogeneously enhancing areas consistent with subareolar
abscess.
On contrast-enhanced images, subareolar abscess showed a benign signal
enhancement pattern (gradual and progressive enhancement without washout).
Nipple inversion appeared as a round, slightly enhanced structure sinking into
the skin, which sometimes resembled a crater. A typical abscess cavity
appeared as a ringlike or irregular structure containing a hypointense
component with an isointense or hyperintense wall
(Fig. 3). Both fistulas and
dilated lactiferous ducts exhibited tubular enhancement; however, these
features differed in location. Fistulas varied in location but definitely led
to a cutaneous orifice. Dilated lactiferous ducts were usually upright
structures leading to the nipple, with or without an abscess cavity at the
other end. If inflammation was present around abscesses or fistulas, it
generally appeared as a disordered slightly enhanced zone.
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —35-year-old woman with subareolar abscess. Contrast-enhanced
T1-weighted image shows abscess cavity (thick arrow) and fistula that
is hypointense linear structure with thin marginal enhancement (thin
arrows).
|
|
Surgery was planned according to clinical findings and radiologic features,
particularly those of high-resolution MRI. Surgical findings such as location
and presence of abscess cavities and fistulas corresponded completely to those
of MRI. Excision of abscesses and fistulas was successfully performed in all
patients. Those with nipple inversion simultaneously underwent reconstruction
of the nipple. No postoperative complications such as bleeding or infection
were observed. All patients were subject to follow-up, which ranged in
duration from 5 to 36 months. No recurrences have been seen to date and
cosmetic appearance has been satisfactory.
Discussion
Subareolar breast abscesses are uncommon but cause prolonged morbidity and
tend to have a chronic, recurring nature. Because the affected patients are
usually young women, the deformation of the nipple and areola that often
accompanies the disease is also distressing.
The pathogenesis of subareolar breast abscess receives little mention in
the literature. Meguid et al.
[2] noted the microscopic
process of squamous metaplasia of major lactiferous ducts in a series of
patients. The primary cause of subareolar breast abscess is the process of
epidermalization or squamous metaplasia of the cuboidal epithelium lining the
ducts and that lining the ampulla, leading to obstruction of the ducts by
keratin plugs. Dilatation of the duct and ampulla occurs because of the
accumulation and stasis of material secreted from the acini. Ultimately,
rupture of the thin columnar epithelial lining of the major duct with
associated bacterial invasion results in the formation of an abscess situated
beneath the areola. Depending on the pathogenesis, appropriate surgical
therapy is warranted including excision of the lactiferous ducts, the affected
ampulla, and distal diseased ducts and fistula (if present) and reconstruction
of the nipple and areola
[2–4].
Preoperatively, it is important to obtain sufficient information about the
lesion. When comparing mammography and sonography, we found high-resolution
MRI to have great advantages in providing a more comprehensive view of the
lesion. The value of mammography was limited by lower sensitivity in this
group of patients; mammography is likely to be negative in young patients with
dense breast parenchyma [5,
6]. Our mammographic results
were similar to those of previous reports. Mammography will, therefore,
continue to play a role in excluding malignancy. Sonography is the most common
imaging option for detecting abscess and guiding abscess drainage
[7]. In our 12 patients, 83%
exhibited significant sonographic findings suggestive of abscess or fistula.
However, for patients with isolated fistulas, particularly if the fistulas
were located inside the nipples and no masses were palpable, sonography failed
to reveal the lesions. The round and irregular surface of the nipple appeared
to be an obstacle and prevented a clear sonographic image of the focus inside
the nipple and ampulla. Furthermore, pressure from the sonography probe can be
too painful for patients with inflammation to tolerate.
View larger version (91K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —30-year-old woman with subareolar abscess. Contrast-enhanced
T1-weighted image shows 1.5-mm fistula inside nipple. Actual fistula cavity is
small hypointense structure and is associated with well-enhanced marginal
inflammatory stroma (arrow).
|
|
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —28-year-old woman with subareolar abscess. Contrast-enhanced
T1-weighted image shows hypointense tubular structure of dilated major
lactiferous duct with enhanced wall (arrows) between inverted nipple
and abscess cavity.
|
|
MRI has recently been reported as a method of assessing breast lesions
[8]. However, MRI with
conventional coils cannot detect lesions less than a few millimeters in size.
Our initial attempt using a standard MRI technique failed in manifesting some
subtle structures of subareolar abscess such as fistula and dilated
lactiferous duct. To overcome this limitation, we used a microscopy coil that
is usually applied in ophthalmology and plastic surgery to allow the
acquisition of images at spatial resolution as high as 100 µm. Because the
microscopy coil is not a special dedicated coil, similar imaging could be
performed with a small surface coil from another manufacturer and the use of
similar signal correction software from a vendor other than the one we
described. Therefore, we believe this type of high-resolution image can be
obtained in most of the MRI centers of average capacity.
Based on our previous study, MR mammary imaging using a microscopy coil can
show ducts measuring 0.8 mm in diameter and lesions as small as 1.0 mm
[9]. In the present study,
high-resolution MRI detected all lesions, even those that could not be
visualized by sonography including a 1.5-mm fistula inside the nipple
(Fig. 4). A complex fistula
with two orifices was displayed as a hartshornlike shape. Moreover, MRI
provided considerably greater detail of lesions than sonography, allowing the
detection of such features as inverted nipples, abscess cavities, fistulas,
dilated lactiferous ducts, and inflammatory signs around the abscesses.
Intraoperative findings confirmed the presence of such features.
When gadodiamide hydrate contrast material was administered, a gradually
and progressively enhanced pattern was evident on T1-weighted images in all
patients, clearly revealing inverted nipples, abscess cavities (visualized as
irregular lesions), and fistulas with enhanced walls and heterogeneous
content. In three cases, the tubular appearance of a dilated major lactiferous
duct was also clearly identified (Fig.
5). Inflammation around the abscesses or fistulas appeared as a
disordered, slightly enhanced zone. This kind of signal enhancement pattern
probably reflects inflammatory cell infiltration, connective tissue
hyperplasia, stroma edema, and vessel regeneration that represent pathologic
changes seen in abscess formation. The present study showed that enhanced MRI
provided more definitive findings than the unenhanced sequences. Further
research should be performed to elucidate different MRI patterns at various
stages of abscess formation.
For achieving a radical cure of the subareolar abscess, adequate surgical
excision is desired. On the other hand, for cosmetic reasons, minimum excision
is preferable. The conventional surgical approach of subareolar breast abscess
was mainly based on the detection of abscess cavities and fistulas using a
fine probe or dye during the operation. Precise preoperative detail was not
easy to obtain, and the incision site was selected by the surgeon empirically.
Especially for those patients who had complex fistulas or surrounding
inflammation, some tiny foci could be neglected. High-resolution MRI with a
microscopy coil provided information regarding the most proper location,
extent, and subtle structure features of the abscess, which made it possible
to design the shortest incision and completely excise the affected area. From
this point of view, we think that high-resolution MRI could change the
surgical approach into a more accurate, thorough, and cosmetic one. Although
high-resolution MRI provided a comparatively comprehensive view of lesions, it
should be noted that surgery was performed with the patient in the supine
position, whereas the MR examinations were performed with the patient prone.
Positional changes should accordingly be taken into account during
preoperative planning.
In conclusion, subareolar breast abscess is a troublesome condition with an
extremely high recurrence rate and therefore deserves increased attention. The
present study showed that high-resolution MRI is a powerful method of
detecting subareolar breast abscess. High-resolution MRI provided a
comprehensive view of inverted nipples, abscess cavities, fistulas, and
inflammation around the lesion. Such information facilitated preoperative
planning. MRI was also able to reveal lesions missed by sonography among
patients without palpable masses. In summary, high-resolution MRI has
potential diagnostic value in subareolar breast abscess. It can be used to
guide surgery with the aim of reducing the recurrence rate.
References
- Zuska JJ, Crile G Jr, Ayres WW. Fistulas of lactiferous ducts.
Am J Surg 1951;81
: 312–317[CrossRef][Medline]
- Meguid MM, Oler A, Numann PJ, Khan S. Pathogenesis-based treatment
of recurring subareolar breast abscesses. Surgery1995; 118:775
–782[CrossRef][Medline]
- Versluijs-Ossewaarde FN, Roumen RM, Goris RJ. Subareolar breast
abscesses: characteristics and results of surgical treatment.
Breast J 2005; 11:179
–182[CrossRef][Medline]
- Yanai A, Hirabayashi S, Ueda K, Okabe K. Treatment of recurrent
subareolar abscess. Ann Plast Surg 1987;18
: 314–318[CrossRef][Medline]
- Kocaoglu M, Somuncu I, Ors F, Bulakbasi N, Tayfun C, Ilkbahar S.
Imaging findings in idiopathic granulomatous mastitis: a review with emphasis
on magnetic resonance imaging. J Comput Assist Tomogr2004; 28:635
–641[CrossRef][Medline]
- Muttarak M. Abscess in the non-lactating breast: radiodiagnostic
aspects. Australas Radiol 1996;40
: 223–225[Medline]
- Hook GW, Ikeda DM. Treatment of breast abscesses with US-guided
percutaneous needle drainage without indwelling catheter placement.
Radiology 1999;213
: 579–582[Abstract/Free Full Text]
- Orel SG, Schnell MD. MR imaging of the breast for the detection,
diagnosis, and staging of breast cancer. Radiology2001; 220:13
–30[Abstract/Free Full Text]
- Kanemaki Y, Kurihara Y, Itoh D, et al. MR mammary ductography using
a microscopy coil for assessment of intraductal lesions.
AJR 2004; 182:1340
–1342[Free Full Text]
- Japan Radiological Society. Mammography
guidelines, 2nd ed. [in Japanese]. Tokyo, Japan: Japan
Radiological Society, 1999
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
