The British Journal of Radiology, 78 (2005), 186188

DOI: 10.1259/bjr/26372381

2005 The British Institute of Radiology

Ultrasound-guided drainage of breast abscesses: results in

151 patients
1

A F CHRISTENSEN, MD, 2N AL-SULIMAN, MD, 1K R NIELSEN, MD, 1I VEJBORG, MD,

N SEVERINSEN, MD, 3H CHRISTENSEN, MD, PhD and 1M B NIELSEN, MD, PhD, DMSc

2
1

Department of Radiology and 2Department of Surgery CE, Rigshospitalet, and 3Department of Neurology, Bispebjerg
Hospital, University of Copenhagen, Denmark

Abstract. The aim of this paper is to describe the efficacy of ultrasound-guided drainage of breast abscesses with
special attention to the risk of recurrence and the need for surgical treatment in a consecutive patient
population. 151 patients, 89 with puerperal and 62 with non-puerperal breast abscesses, were treated with
ultrasound-guided drainage, by needle or catheter under local anaesthesia. Follow-up punctures were performed
at 2 or 3 day intervals until the clinical condition and ultrasound findings had improved. All patients were
treated with oral antibiotics. Mammography was performed to search for underlying cancer. 86 (97%) out of 89
patients with puerperal abscesses and 50 (81%) out of 62 with non-puerperal abscesses recovered after the first
round of ultrasound-guided drainage. One patient in each group had recurrence in loco but recovered after
further ultrasound-guided drainage. 13 patients, 11 with non-puerperal and two with puerperal abscesses,
underwent surgical excision of the abscess cavity or fistulas. Breastfeeding continued and 117 patients were
treated as outpatients. The median number of follow up examinations in the ultrasound-department was four
(range 110) for the group of patients with puerperal abscess and three (range 17) in the group of patients with
non-puerperal abscess. The corresponding figure for the median number of punctures was for both groups one
(range 16 and 14). There were no reports of newly diagnosed breast cancer in the 2 year follow-up period.
This study supports the use of ultrasound-guided drainage in puerperal and non-puerperal breast abscesses. The
method is less invasive than traditional surgery and has a high rate of success.

Ultrasound-guided drainage has been described in

smaller series as a treatment for breast abscesses, the
treatment for which has traditionally been surgery or less
frequently puncture without imaging guidance [15].
Ultrasound guided drainage is now widely accepted as
first-line treatment. The aim of this study was to describe
the efficacy of ultrasound guided drainage with special
attention to resolution and to the risk of recurrence as well
as the need for surgical treatment in a consecutive patient
population.

Materials and methods

From 1 January 1998 to 31 December 2001, 248 patients
were referred to the ultrasound department with clinical
suspicion of new-developed breast abscess. In 87 (35%) of
these patients breast abscess was excluded by ultrasound
examination. The remaining 161 patients had a breast
abscess. 10 of these 161 patients were not treated by
ultrasound-guided drainage: five patients had abscesses
which perforated spontaneously before treatment, three
patients underwent surgery because the abscesses were
too superficial and not suited for ultrasound-drainage
with skin changes which required surgery for healing,
one patient refused puncture under local anaesthesia
and one 87-year-old patient had syncope before treatment was initiated and no further treatment was undertaken. The remaining 151 patients were treated with
Received 22 March 2004 and in revised form 27 August 2004, accepted
4 October 2004.

186

ultrasound-guided drainage of an abscess. They form the

basis of the study.
The 151 patients that underwent ultrasound-guided
drainage were all female, 89 had puerperal breast abscesses
(median age 32 years, range 1542 years), and 62 had nonpuerperal abscesses (median age 42 years, range 1572
years). None of the patients were undergoing treatment for
other breast pathology.
Rigshospitalet is the centre for breast disease in central
Copenhagen, covering a population of approximately
600 000 individuals. All patients with breast disease are
referred to the clinic for breast surgery. If breast abscess is
clinically suspected patients are referred to ultrasound for
diagnosis and if possible for ultrasound-guided drainage
with either a 0.81.2 mm needle or, if the cavity is larger
than 3 cm, a 5.7 F (2 mm) one-step catheter. All punctures
are performed under local anaesthesia. The procedure can
be performed at any time in our institution when clinically
indicated.
Follow-up ultrasound examinations are performed at
23 day intervals and if the abscess is still present, another
puncture is performed or the catheter is left in place.
Ultrasound follow-up ends when the clinical condition
improves and two consecutive examinations have shown
no residual abscess. The final examination is performed
after the final puncture or removal of the catheter in the
examination before. The catheter is cared for by the
patient herself after instruction is given on the ward.
The removal of a catheter requires clinical improvement of
the local symptoms and no residual undrained fluid in the
abscess visualized by ultrasound. All patients received oral
antibiotic treatment; dicloxacillin 1 g three times a day in
The British Journal of Radiology, March 2005

Ultrasound-guided drainage of breast abscesses

all patients and additionally metronidazole 0.5 g three

times a day for non-puerperal abscesses. Patients allergic
to penicillin receive erythromycin 500 mg three times a
day. All 151 patients in this study followed this procedure.
In 2003 and 2004 the hospital case records were
retrospectively reviewed. The following parameters were
registered: patient data, details about drainage procedure
and punctures, bacteriology and surgical data and possible
recurrent abscess within 24 months. Clinical examination
and mammography, ultrasound and where indicated
ultrasound-guided biopsy, were performed within 3
months after the resolution of the abscess, mainly in the
group of patients with non-puerperal breast abscesses. In
January 2004 all 151 patients were followed in the register
of the Danish Breast Cancer Group to see if any had been
diagnosed with breast cancer within 24 months.
Statistical analyses were performed with SPSS for
Windows 9.0 (SPSS Inc., Chicago, IL). Descriptives
included numbers, percentage and range. Logistic regression analysis was first performed by univariate testing,
where variables were tested based on literature and clinical
experience. We planned to include variables reaching a
significance level of ,0.1 in univariate testing in a final
multivariate logistic regression model. The outcome
variable used was resolution after ultrasound-guided
drainage, being defined as no recurrent abscess and no
need for surgery.

Results
Patient data are given in Tables 1 and 2. 136 patients, 86
(97%) of the 89 patients with puerperal breast abscess and
50 (81%) of the 62 patients with non-puerperal breast
abscess had no signs of remaining infection in the breasts
after the first ultrasound-guided drainage treatment. 77
(87%) of the patients with puerperal breast abscess were
treated as outpatients despite catheter drainage treatment.
The catheters were cared for by the patients themselves

after instruction and removed at the hospital after

(median) 4 days (range 26 days).
After successful ultrasound-guided treatment of the
abscess, three of these patients underwent surgery based
solely upon an unconfirmed suspicion of cancer: the
findings of a solid nodule in one patient, atypical cells in a
biopsy in another patient and an abscess in an atypical
cyst.
Two patients, one patient in each group had a recurrent
abscess (after 5 months and 10 months, respectively) but
recovered after further ultrasound-guided drainage.
13 patients, two with puerperal and 11 with nonpuerperal abscesses, underwent surgical excision of the
abscess cavity. Two patients preferred surgery to ultrasoundguided drainage, one had pain after the abscess resolved
and two other patients had an abscess that perforated
during treatment. In six cases surgery was for a complicating fistula, one puerperal and five non-puerperal
two of the non-puerperal fistulae were present prior to
commencement of treatment. One patient was not
effectively treated with ultrasound-drainage, and one had
pus in ectatic ducts and was difficult to treat with
ultrasound-guided drainage.
The median number of follow up examinations in the
ultrasound department was four (range 110) for the
group of patients with puerperal abscess and three (range
17) in the group of patients with non-puerperal abscess.
The corresponding figure for the median number of
punctures was for both groups one (range 16 and 14).
The two patients who had the most extensive treatment
without reference to surgery needed six punctures and 10
ultrasound examinations each. Both patients had continuous improvement in both clinical condition and ultrasound findings which was why surgery was not performed.
The results of the univariate logistic regression analysis
are presented in Table 3. Recovery was seen in 135 out of
the 151 patients (87%). In this analysis only a puerperal
history reached significance. A multivariate analysis was
therefore not performed.

Table 1. Clinical characteristics of 151 patients with breast abscesses

No. of patients
Age (years)
Days from symptom onset to diagnosis and treatment
Needle puncture/catheter drainage
Hospital admittance/outpatients
Median no. of follow-up ultrasound examinations (range)
Central vs peripheral localization of abscess
Median size of abscess (in cm) (range)
No. of punctures

Puerperal abscess

Non-puerperal abscess

89
32 (1542)
7 (270)
23 (26%)/66 (74%)
12 (13%)/77 (87%)
4 (110)
34.5%/65.5%
3.5 (1.08.0)
1 (16)

62
42 (1572)
7 (130)
46 (74%)/16 (26%)
22 (36%)/40 (64%)
3 (17)
43.5%/56.5%
2.0 (0.58.0)
1 (14)

Data are given as median (range) except where indicated otherwise.

Table 2. The outcome of the treatment in the 151 patients

Outcome of the treatment

Recovered after the first round of ultrasound-guided drainage

Had recurrent abscess after end ultrasound-guided drainage
Underwent surgery (see results for details)

The British Journal of Radiology, March 2005

Breast abscesses
Puerperal no. of patients

Non-puerperal no. of patients

86 (97%)
1 (1%)
2 (2%)

50 (81%)
1 (1%)
11 (18%)

187

A F Christensen, N Al-Suliman, K R Nielsen et al

Table 3. Univariate logistic regression analyses of factors possibly affecting breast abscess resolution in 151 females treated
with ultrasound-guided drainage of breast abscesses

Puerperal vs non-puerperal
Age+10 years
Delay from symptom onset to treatment
Total number of abscesses
Largest diameter+10 mm
Central vs peripheral localization
Catheter drainage vs only puncture

OR CI 95%

p-value

4.1
0.8
1.0
0.9
1.6
1.5
0.6

0.0471
0.504
0.6531
0.896
0.175
0.477
0.515

1.0216.7
0.51.4
0.91.1
0.16.8
0.83.0
0.54.1
0.12.9

OR, odds ratio; CI, confidence interval.

26 patients with puerperal and 51 patients with nonpuerperal abscess were referred for clinical evaluation and
mammography. All of the 26 patients with puerperal
abscess had palpation findings as the basis for referral for
mammography. The patients with non-puerperal abscess
were referred for mammography unless the abscess was
situated centrally and subareolar (only three patients with
lateral abscesses were not referred). In the follow-up
period there were no reports of cases of de novo diagnosed
breast cancer in the two patient groups.

Discussion
Ultrasound-guided drainage in combination with oral
antibiotics was shown to be an effective treatment for
breast abscess, especially for the group with puerperal
abscesses. No other factors, including whether treatment
was conducted using needle puncture and aspiration or
catheter drainage, had any independent effect on the recovery rate.
The high rates of recovery found in this study confirm
previous studies, where recovery rates exceeding 90% using
ultrasound-guided drainage have been reported [1, 46].
To our knowledge no previous study has included more
than 30 patients. Treatment failure using ultrasoundguided drainage has been reported in cases where either
the abscess has been larger than 3 cm in diameter or it
was placed centrally in a subareolar position [1]. This
study did not support these findings since neither size nor
localization showed any independent effect on the recovery
rate.
In some institutions the standard treatment still remains
early incision under general anaesthesia combined with
drainage tube insertion [7]. The side effects of this
treatment include scarring and termination of breastfeeding [3]. Ultrasound-guided drainage causes less scarring, does not affect breast-feeding and does not require
general anaesthesia or hospitalization [3]. Ultrasoundguided drainage is a less expensive procedure than surgery
[9]. Based on our findings, ultrasound-guided drainage
treatment should replace surgery as the first line of
treatment in uncomplicated puerperal or non-puerperal
breast abscess. Furthermore this technique is not difficult to master and the service can be provided on a 24 h
basis.

188

In treatment resistant cases, where the abscess is

unresponsive to the combination of repeated drainage
and oral antibiotics, surgical treatment still has a role. It is
unclear why some abscesses are unresponsive to treatment.
Surgery is also needed in cases with superficial abscesses
with skin changes. This can make a surgical excision
necessary for healing. Surgery can also be necessary in
special cases where other concerns such as suspicion of
malignancy or the need for focus excision are important
factors. Carcinoma of the breast may be confused with
inflammatory conditions and thus pose a diagnostic
problem [8]. We use early follow-up mammography
within 3 months and when appropriate ultrasoundguided biopsy to exclude this diagnosis. In this study no
patients had diagnosed breast cancer in the follow-up
period. The number of patients included in this study is in
our opinion not large enough to substantiate any
statement about whether clinical mammography should
be routinely undertaken. Likewise we have not found any
reason to alter indications for mammography.
Needle puncture of lactating breasts has previously been
associated with fistula formation [10]. In this study, six
patients had a fistula in connection with a breast abscess.
However only four of the 151 patients receiving ultrasound-guided drainage developed a fistula, of which only
one had a puerperal abscess. Two of the six patients had a
fistula before treatment started. These data indicate that
ultrasound-guided drainage does not cause more fistula
formation in puerperal than in non-puerperal abscesses.
In conclusion, this study supports the use of ultrasoundguided drainage in both puerperal and non-puerperal
breast abscesses. The method is less invasive than
traditional surgery and has a high rate of success.

References
1. Hook GW, Ikeda DM. Treatment of breast abscesses with
US-guided percutaneous needle drainage without indwelling
catheter placement. Radiology 1999;213:57982.
2. Tan SM, Low SC. Non-operative treatment of breast
abscesses. Aust N Z J Surg 1998;68:4234.
3. Schwarz RJ, Shrestha R. Needle aspiration of breast
abscesses. Am J Surg 2001;182:1179.
4. Karstrup S, Solvig J, Nolsoe CP, et al. Acute puerperal breast
abscesses: US-guided drainage. Radiology 1993;188:8079.
5. Blaivas M. Ultrasound-guided breast abscess aspiration in a
difficult case. Acad Emerg Med 2001;8:398401.
6. Karstrup S, Nolsoe C, Brabrand K, Nielsen KR.
Ultrasonically guided percutaneous drainage of breast
abscesses. Acta Radiol 1990;31:1579.
7. Marchant DJ. Inflammation of the breast. Obstet Gynecol
Clin North Am 2002;29:89102.
8. Das DK, Sodhani P, Kashyap V, Parkash S, Pant JN,
Bhatnagar P. Inflammatory lesions of the breast: diagnosis by
fine needle aspiration. Cytopathology 1992;3:2819.
9. Imperiale A, Zandriono F, Calabrese M, Parodi G, Massa T.
Abscesses of the breast: US-guided serial percutaneous
aspiration and local antibiotic therapy after unsuccessful
systemic antibiotic therapy. Acta Radiologica 2001;42:161.
10. Barker P. Milk fistula. An unusual complication of breast
biopsy. J RC Surg Edinburgh 1988;106.

The British Journal of Radiology, March 2005