The British Journal of Radiology, 81 (2008), 743748

SHORT COMMUNICATION

Whole-body PET/CT-mammography for staging breast cancer:

initial results
1
T-A HEUSNER, MD, 2L S FREUDENBERG, MD, 1H KUEHL, MD, 1E A M HAUTH, MD, 1P VEIT-HAIBACH, MD,
1
M FORSTING, MD, 2A BOCKISCH, MD, PhD and 1G ANTOCH, MD

Departments of 1Diagnostic and Interventional Radiology and Neuroradiology and 2Nuclear Medicine, University
Hospital Essen, Germany

ABSTRACT. The purpose of this study was to evaluate the feasibility and utility of a
dedicated positron emission tomography (PET)/CT protocol in breast cancer patients. 40
patients with suspected recurrent breast cancer underwent whole-body PET/CT in the
supine position (SP) followed by PET/CT of the breasts and axillae in the prone position
(PP) using a special positioning aid. PP and SP images were compared in terms of the
tumour-to-thoracic-wall distance, tumour-to-skin distance and tumour volume,
diameter, density, maximal standardized uptake value (SUVmax) and localization. The
size of axillary areas, the number of intra-axillary lymph nodes, their transverse
diameters, their SUVmax and the number of distant metastases were compared between
PP and SP images. Differences were tested for significance using the Students t-test. All
patients tolerated PP imaging well. Five locally recurrent breast cancers were detected,
both in the SP and in the PP. Mean tumour-to-thoracic-wall distances (PP, 19 mm; SP,
8 mm; p50.003) and tumour-to-skin distances (PP, 10 mm; SP, 7 mm; p50.013) were
significantly larger in the PP than in the SP. Potential thoracic wall or skin infiltration, as
well as quadrant localization, were determined more easily in PP. The axillary area was Received 9 June 2007
wider in the PP when compared with SP (PP, 14.4 cm2; SP, 10.6 cm2; p,0.001). No other Revised 1 August 2007
parameters were significantly different. In conclusion, a dedicated whole-body PET/CT Accepted 29 August 2007
examination, including PET/CT mammography, is feasible for clinical practice and may
DOI: 10.1259/bjr/69647413
offer important information on the possible infiltration of a breast lesion into the
adjacent thoracic wall and skin. Even though the axilla may be delineated more clearly 2008 The British Institute of
in the PP, there seems to be no benefit with regard to N-staging. Radiology

Accurate tumour staging represents a precondition in
breast cancer patients to determine further therapy.
Different imaging modalities for tumour staging are
available; however, X-ray mammography is the most
widely used technique for diagnosis of the primary
lesion in both symptomatic and asymptomatic patients
[1, 2]. Correlation of mammography findings with those
from ultrasound and MRI has been found to be helpful
for the differential diagnosis of a breast lesion and for the
detection of occult breast tumours [3, 4]. Mammography
is complemented by staging for locoregional lymph node
metastases and distant metastases. This multimodality,
multistep staging algorithm may include chest radio-
graph, ultrasound, CT of the chest and/or abdomen and
tumour markers.
The growing availability of dual-modality PET/CT
systems opens new diagnostic oncological strategies [5,
6]. PET/CT has been found to be beneficial in patients
with breast cancer when compared with conventional
Address correspondence to: Gerald Antoch, Department of
Diagnostic and Interventional Radiology and Neuroradiology,
University Hospital Essen, University at Duisburg-Essen,
Germany. E-mail: gerald.antoch@uni-duisburg-essen.de
staging algorithms [7, 8]. However, PET/CT has been
used primarily for the assessment of potential lymph
node metastases and distant metastases in breast cancer
patients. Mammography in conjunction with ultrasound,
as well as MR mammography, has remained the method
of choice for imaging the breast and primary tumour.
Theoretically, patient positioning similar to that per-
formed in MR mammography may provide more
accurate information on the primary tumour and axilla
in PET/CT. This study was designed to assess the
technical feasibility of a disease-defined PET/CT proto-
col for breast cancer patients that combined whole-body
PET/CT staging with PET/CT mammography.
Methods and materials/patients
40 female patients (mean age, 58 years11 years) who
had received initial treatment for breast cancer but who
had suspected disease progression and/or suspected
recurrent breast lesions (based on rising tumour markers,
clinical findings or mammography findings) were
included. Histopathology of recurrent breast lesions
was available for image correlation. Histopathology of

The British Journal of Radiology, September 2008 743


metastases was not available. This retrospective study
was performed in accordance with the regulations of the
local ethics committee. Written consent was obtained
from every patient concerning PET/CT imaging.
PET/CT system/workstation
PET/CT imaging was performed on a biograph2 PET/
CT system (Siemens Molecular Imaging, Hoffman Estates,
IL) composed of a dual-slice CT scanner (Somatom
Emotion2; Siemens Medical Solutions, Forchheim,
Germany) and a full-ring PET scanner (ECAT HR+2;
Siemens Molecular Imaging, Hoffman Estates, IL). The
PET system had an axial field of view of 15.5 cm per bed
position and an in-plane spatial resolution of 4.6 mm. CT
was performed first, followed by PET.
All measurements were performed on an AW Suite2
Workstation (General Electrics Healthcare, Munich,
Germany). Differences between measurements made in
the supine position (SP) and the prone position (PP) were
tested for significance (p,0.05) with the Students t-test.
One radiologist and one nuclear medicine physician with
more than 3 years of experience in PET/CT evaluated all
images in consensus.
Imaging protocol
The dedicated breast PET/CT protocol consisted of
two parts. Firstly, a whole-body PET/CT scan was
performed in the SP, covering a field of view from the
head to the upper thighs. Image acquisition was
performed in the caudocranial direction with 100 mAs
and 130 kV. 140 ml of an intravenous contrast agent
(Ultravist 3002; Schering AG, Berlin, Germany) contain-
ing 300 mg ml1 of iodine were administered with an
automated injector (XD 55002; Ulrich Medical Systems,
Ulm, Germany) with a flow rate of 3 ml s1 for the first
90 ml, and 1.5 ml s1 for the following 50 ml. The start
delay was 50 s. Images were reconstructed with a 5 mm
slice thickness and a 2.4 mm increment. Following
acquisition of the CT data, PET images were obtained
60 min after injection of ,340 MBq of 18F-fluorodeoxy-
glucose (FDG). PET emission time was adapted to the
patients body weight: ,65 kg, 4 min per bed position;
6585 kg, 5 min per bed position; and .85 kg, 6 min per
bed position. Iterative algorithms (FORE (Fourier rebin-
ning) and AWOSEM (attenuation-weighted expectation
maximization), non-linear) with two iterations and eight
subsets were used for image reconstruction. Data were
filtered (FWHM (full width at half maximum) 5.0 mm)
and scatter was corrected.
The second part of the breast-specific protocol was
performed after repositioning the patient into the PP
using a special breast positioning aid (Additec Mamma
Comfort2; Additec GmbH, Markt Indersdorf, Germany;
Figure 1). A topogram in the lateral view was performed
to define the scan range from the axilla to the lower end of
the breasts. No additional contrast medium was applied
for PET/CT in the PP. Image acquisition was performed
in a caudocranial direction. CT parameters were the same
as those in the SP. The number of PET bed positions was
adapted to include the breasts and axillae. This resulted in
744
T-A Heusner, L S Freudenberg, H Kuehl et al
Figure 1. Breast positioning device (Additec Mamma
Comfort2) made from foam plastic. The device is constructed
for prone breast positioning.
one or two bed positions for the prone scan. PET emission
time was set to either 6 min or 7 min, depending on the
volume of the breast. PET image reconstruction was
performed according to the SP protocol.
Technical feasibility
All patients were questioned for potential discomfort
during prone imaging compared with supine imaging.
The additional time (min) required for prone imaging
was reported, as was the number of additionally
required bed positions.
Delineation of breast lesions
In the case of local recurrence, the lesion was localized
within a specific quadrant and these results were
compared with operative results and histopathology.
Lesions were identified by elevated focal tracer uptake
on PET/CT. Size measurements were taken from CT
images using the distance and volume measuring func-
tions of the AW Suite2 5.5.3e Volume Viewer Plus2
Workstation (General Electrics Healthcare). Tumour-to-
skin and tumour-to-thoracic-wall distances were deter-
mined (mm) to assess any potential infiltration of the
lesion into adjacent structures (Figure 2). The largest axial
diameters and vertical diameters of all lesions were
measured (mm), and the lesion volumes (mm3) were
assessed. Maximal standardized uptake values (SUVmax)
and mean densities (Hounsfield Units (HU)) of all breast
lesions were measured in both the PP and the SP.
Assessment of ipsilateral axilla
Axillary regions were evaluated for clear anatomical
depiction in the SP and the PP; the area of axillary fat (i.e.
the area between the outer margin of the latissimus
dorsi/major teres muscle and the minor/major pector-
alis muscle) was measured (mm2) (Figure 3). From PP
and SP imaging, the number of detectable intra-axillary
lymph nodes was recorded, and the transverse diameters
(mm) of all intra-axillary lymph nodes were measured.
In addition, the SUVmax of all axillary lymph nodes with
qualitatively detectable increased tracer uptake was
measured on both the supine and the prone scans.
The British Journal of Radiology, September 2008
Short communication: Whole-body PET/CT-mammography for staging breast cancer

(a) (b)

(c) (d)

Figure 2. Contralateral breast cancer manifestation in a 47-year-old woman on (a,b) fluorodeoxyglucose positron emission
tomography (FDG-PET)/CT and on (c,d) CT; the tumour can be more clearly distinguished from adjacent structures in the prone
position (a,c) than in the supine position (b,d). On prone imaging, thoracic wall infiltration can be clearly negated because of
the fatty tissue separating the tumour and thoracic wall in the prone position (c) but not in the supine position (d).

Assessment of distant metastases
The number of distant metastases within the field of
view of SP and PP protocols was reported.
This included the repositioning of the patient and the PP
PET/CT investigation. The mean PET emission time was
6.50.5 min per bed position in the PP. In 28 patients, 1
bed position was scanned in the PP, whereas 2 bed
positions were scanned in the PP in 12 patients.

Results
Technical feasibility
All patients tolerated PET/CT in the PP well. The
additional time required for PP PET/CT was 205 min.
Delineation of breast lesions
6 of the 40 patients suffered from histopathologically
confirmed local breast cancer recurrence (mean age,
50.4 years; range, 4266 years; standard deviation (SD),

The British Journal of Radiology, September 2008 745


(a)
Figure 3. Visualization of axillary fat in (a) the prone position and (b) the supine position on CT. The axillary area measures
40 cm2 in the prone position and 24 cm2 in the supine position. Prone positioning offers a more extensive evaluation of the
axillary fat and its lymph nodes. 1, major pectoral muscle; 2, minor pectoral muscle; 3, latissimus dorsi muscle; 4, major teres
muscle.
9.7 years). Of these six local recurrences, five were
detected by PET/CT mammography. One small FDG-
PET-negative multifocal tumour was not identified by
either the PET or the CT protocol. The histological sizes
of the three small tumour lesions constituting this
moderately differentiated, multifocal, invasive ductal
mammarian cancer were 11 mm, 5 mm and 5 mm in
diameter. Two ipsilateral and three contralateral recur-
rences were identified with PET/CT in both the SP and
the PP. In one breast lesion, quadrant localization was
impossible in the SP but was achieved in the PP (lower
outer quadrant). The mean tumour-to-skin distances (SP,
78 mm; PP, 1013 mm; p50.013) and mean tumour-
to-thoracic-wall distances (SP, 817 mm; PP,
1931 mm; p50.003) were significantly higher in the
PP than in the SP (Figure 2), indicating better delineation
of the tumour from the thoracic wall and the skin. No
significant differences were detected between the PP and
the SP when assessing the maximal axial lesion diameter
(SP, 5543 mm; PP, 5739 mm; p50.465), the maximal
vertical lesion diameter (SP, 2011 mm; PP, 1911 mm;
p50.189), lesion volume (SP, 8.78.4 cm3; PP,
9.28.6 cm3; p50.119), lesion SUVmax (SP, 6.98.9; PP,
7.410.0; p50.396) or mean lesion density (SP,
4122 HU; PP, 4423 HU; p50.487). The sizes of the
Table 1. Sizes of breast cancer lesions detected with PET/CT
Patient no. Size in supine position (mm)
1 21613
2 63635
3 12568
4 49626
5 19616
746
T-A Heusner, L S Freudenberg, H Kuehl et al
(b)
different breast cancer lesions detected with PET/CT are
shown in Table 1.
Assessment of the ipsilateral axilla
Because of the significantly wider area of axillary fat
seen in the PP (14.47.3 cm2) compared with the SP
(10.64.7 cm2; p,0.001), different anatomical structures
of the axilla may be more easily differentiated from each
other in the PP (Figure 3). No significant differences
were detected in the number of lymph nodes (n587 each
for PP and SP; p51.0), their transverse diameters (SP,
52 mm; PP, 52 mm; p50.915) or the SUVmax of
lymph nodes with qualitatively increased tracer uptake
(SP, 6.53.2; PP, 5.52.8; p5 0.061).
Assessment of distant metastases
40 distant metastases were detected within the limited
field of view of the prone protocol in 16 patients. When
assessing the same field of view in the SP, the same
number of distant metastases was detected. The location
of the distant metastases is shown in Table 2.
Size in prone position (mm)
30613
64633
11965
51626
19616
The British Journal of Radiology, September 2008
Short communication: Whole-body PET/CT-mammography for staging breast cancer
Table 2. Location of the distant metastases
Localization Number of distant metastases detected
with PET/CT
Bone 15
Lymph nodes 12
Lung 10
Pleura 2 2
Liver 1 2
All lymph nodes listed are distant lymph node metastases, e.g. hilar or infracarinal lymph node metastases. 4 of 15 bone
metastases (26.7%) were not detected with CT alone, as were 3 of 12 lymph node metastases (25%).
PET, positron emission tomography.
Discussion
Initial results indicate that prone breast positioning
may improve the assessment of any potential infiltration
of breast tumour into the thoracic wall or the skin on
PET/CT. In addition, the axilla may be assessed more
easily for potential metastatic spread because of
enhanced anatomical visualization. Therefore, breast
imaging in the PP may be a helpful adjunct to whole-
body PET/CT staging. Further studies are required to
assess the accuracy of whole-body PET/CT mammogra-
phy for TNM (tumour, node, metastases) breast cancer
staging.
Different quadrants of the breast can be distinguished
more easily in the PP, with the potential to more
accurately localize a breast tumour. In particular, the
axillary tail of the breast seemed to be visualized more
thoroughly in the PP. This may be of particular interest
because the axillary tail of the breast harbours 48% of all
breast cancer manifestations [9]. This advantage of PP
compared with SP has been demonstrated for MR
mammography, where prone imaging is the method of
choice for visualization of the breast [10]. To improve
PET accuracy when assessing the breast for potential
lesions, breast positioning similar to that of MRI has been
proposed for PET imaging [11]. A substantial advantage
of the PP is the potential to better differentiate the
tumour from its adjacent structures. In the PP, fatty
tissue, as well as glandular tissue of the breast, is
uncompressed, thus offering clearer visualization of the
fatty lamella that separates the tumour from the thoracic
wall or the skin. In cases of thoracic wall or skin
infiltration, this fatty lamella will disappear, indicating
tumour invasion. Infiltration of the pectoral muscles or
the skin owing to a T4 carcinoma of the breast has
implications for patient management [12]. Thus, the early
detection of such locally advanced disease must be
considered of interest even before it has been detected by
pathology.
MR mammography has been found to have a higher
sensitivity for detecting malignant breast lesions than
FDG-PET [13]. Indeed, FDG-PET is at a disadvantage
when detecting malignant breast tumour lesions ,1 cm
because of its limited spatial resolution and the low
glucose uptake of well-differentiated tumours [2].
However, compared with other functional imaging
modalities, FDG-PET is still considered the most sensi-
tive functional method for the detection of primary
breast carcinoma [2]. A sensitivity of 88% for the
detection of malignant breast lesions has been documen-
ted [14]. Anatomical correlation for FDG-PET, as
The British Journal of Radiology, September 2008
Number of metastases occult on CT
alone
4
3
2
provided by PET/CT, may offer an even more thorough
evaluation of breast lesions compared with PET alone.
Further studies will have to assess the accuracy of this
imaging protocol in clinical practice to determine if its
diagnostic accuracy can compete with MR mammogra-
phy.
Kumar et al [15] reported a significant increase of
12.6% (SUVmax) in dual-point measurements of breast
cancer lesions over time with FDG-PET. This finding
differed from inflammatory lesions and normal breast
tissue, for which the SUVmax decreased over time. In
agreement with Kumar et al [15], an increase in the
SUVmax of 9.7% was detected between the SP and the PP
in this study. In this setting, the SP and PP may be
considered dual-point measurements. However, based
on the small number of lesions in the current data
analysis, further studies are required to assess whether a
whole-body protocol including breast imaging in the SP
and PP will improve the differentiation of malignant and
benign lesions.
The application of additional iodinated contrast
material for the prone imaging method has to be
discussed. Potentially, this may further improve image
quality. In a study by Boone et al [16] the authors
emphasise the importance of intravenous contrast for
lesion detection in CT mammography. In lesions with
low FDG uptake, lesion detection might be improved if
applying contrast material not only for supine whole-
body staging PET/CT but also for the prone breast PET/
CT.
A whole-body PET/CT protocol in the PP instead of
the combined whole-body SP/PP protocol should also be
discussed, as this would substantially reduce the
examination time while offering good breast lesion
delineation owing to the PP of the patient. However,
patient tolerance may limit this approach, as the patient
would have to be prone for approximately 30 min.
However, daily routine has demonstrated good tolerance
of prone breast MRI, which requires similar examination
times. The use of 64-row multislice PET/CT systems will
further reduce the examination times compared with
PET/CT scanners with fewer detector rows. In this
setting, a whole-body prone protocol may be clinically
feasible.
Conclusions
Whole-body PET/CT mammography is technically
feasible in clinical practice. If whole-body PET/CT is
indicated for tumour staging in patients with suspected
747

breast cancer recurrence, PP imaging of the breasts may
add important information on tumour infiltration com-
pared with SP imaging alone. Even though the axilla
may be delineated more clearly in the PP, there seems to
be no benefit with regard to N-staging compared with SP
imaging. Owing to the small sample size of this initial
report and the rather large mean breast tumour size,
further studies will need to assess the actual accuracy of
this combined breast staging protocol compared (i) with
conventional PET/CT imaging and (ii) with a combina-
tion of PET/CT with mammography or MR mammo-
graphy.
Acknowledgments
We thank Thomas Beyer, PhD, for his organizational
support and Additec GmbH for providing the position-
ing aid.
References
1. Agnese DM. Advances in breast imaging. Surg Technol Int
2005;14:516.
2. Scheidhauer K, Walter C, Seemann MD. FDG PET and other
imaging modalities in the primary diagnosis of suspicious
breast lesions. Eur J Nucl Med Mol Imaging 2004;31:709.
3. Buscombe JR, Holloway B, Roche N, Bombardieri E.
Position of nuclear medicine modalities in the diagnostic
work-up of breast cancer. Q J Nucl Med Mol Imaging
2004;48:10918.
4. Rausch DR, Hendrick RE. How to optimize clinical breast
MR imaging practices and techniques on your 1.5-T system.
Radiographics 2006;26:146984.
5. Antoch G, Vogt FM, Freudenberg LS, Nazaradeh F, Goehde
SC, Barkhausen J, et al. Whole-body dual-modality PET/CT
and whole-body MRI for tumor staging in oncology. JAMA
2003;24;290:3199206.
748
T-A Heusner, L S Freudenberg, H Kuehl et al
6. Czernin J, Allen-Auerbach M, Schelbert HR. Improvements
in cancer staging with PET/CT: literature-based evidence
as of September 2006. J Nucl Med 2007;48:7888.
7. Zangheri B, Messa C, Picchio M, Gianolli L, Landoni C,
Fazio F. PET/CT and breast cancer. Eur J Nucl Med Mol
Imaging 2004;31:13542.
8. Tatsumi M, Cohade C, Mourtzikos KA, Fishman EK, Wahl
RL. Initial experience with FDG-PET/CT in the evaluation
of breast cancer. Eur J Nucl Med Mol Imaging 2006;33:-
25462.
9. Pfleiderer A, Breckwoldt M, Martius G. Invasives
Mammkarzinom. In: Pfleiderer A, editor. Gynakologie
und Geburtshilfe. Stuttgart: Thieme Verlag; 2002:219.
10. El Yousef SJ, Duchesneau RH, Alfidi RJ, Haaga JR, Bryan PJ,
LiPuma JP. Magnetic resonance imaging of the breast.
Radiology 1984;150:7616.
11. Brix G, Henze M, Doll J, Lucht R, Zaers J, Trojan H, et al.
Diagnostic evaluation of the breast using PET: optimization
of data aquisition and postprocessing. Nuklearmedizin
2000;39:626.
12. Pfleiderer A, Breckwoldt M, Martius G. Invasives
Mammkarzinom: Therapie. In: Pfleiderer A, editor.
Gynakologie und Geburtshilfe. Stuttgart, New York:
Thieme Verlag; 2002:224.
13. Walter C, Scheidhauer K, Scharl A, Goering UJ, Theissen P,
Kugel H, et al. Clinical and diagnostic value of preoperative
MR mammography and FDG-PET in suspicious breast
lesions. Eur Radiol 2003;13:16516.
14. Samson DJ, Flamm CR, Pisano ED, Aronson N. Should FDG
PET be used to decide whether a patient with an abnormal
mammogram or breast finding at physical examination
should undergo biopsy? Acad Radiol 2002;9:77383.
15. Kumar R, Loving VA, Chauhan A, Zhuang H, Mitchell S,
Alavi A. Potential of dual-time-point imaging to improve
breast cancer diagnosis with (18)F-FDG PET. J Nucl Med
2005;46:181924.
16. Boone JM, Kwan ALC, Yang K, Burkett GW, Lindfors KK,
Nelson TR. Computed tomography for imaging the breast.
J Mammary Gland Biol Neoplasia 2006;11:10311.
The British Journal of Radiology, September 2008