| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Dipartimento di Scienze Biomorfologiche e Funzionali, Università degli
Studi di Napoli Federico II, Via Posillipo 196, 80123, Naples, Italy.
2 National Research Council Napoli, Institute of Bio-structure and Bio-imaging,
SDN Foundation, Naples, Italy.
3 Dipartimento di Chirurgia, Università degli Studi di Napoli Federico
II, Naples, Italy.
Received November 18, 2007;
accepted after revision May 14, 2008.
Address correspondence to M. Imbriaco
(mimbriaco{at}hotmail.com).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty-four patients with 55 breast lesions
underwent two PET/CT scans (dual-time-point imaging) in the prone position and
breast MRI. Sensitivity, specificity, and overall accuracy were calculated. In
addition, the average percentage of change in standard uptake values
(
%SUVmax) between time point 1 and time point 2 was
calculated for PET/CT. A final histopathologic diagnosis was available for all
patients.
RESULTS. MRI showed an overall accuracy of 95%, with sensitivity and
specificity of 98% and 80%. Conversely, dual-time-point PET/CT showed an
accuracy of 84% for lesions with an SUVmax 
2.5 or with a
positive %SUVmax, with sensitivity and specificity of 80%
and 100% versus 69% accuracy, 62% sensitivity (both, p < 0.001),
and 100% specificity (p not significant) for single-time-point
PET/CT. On PET/CT, malignant lesions showed an increase in FDG between time
points 1 and 2, with a %SUVmax of 11 ± 24. Benign
lesions showed either no change or a decrease in SUVmax between
time points 1 and 2, with a %SUVmax of –21 ±
7.
CONCLUSION. A dual time point improves PET/CT accuracy in patients with a suspected breast malignancy over single-time-point PET/CT. On PET/CT, FDG is increasingly taken up over time in breast tumors; conversely, benign lesions show a decrease in FDG uptake over time. These changes in SUV might represent a reliable parameter that can be used to differentiate benign from malignant lesions of the breast on PET/CT examination.
Keywords: breast carcinoma • FDG PET • MRI • PET/CT
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fluorine-18 FDG PET and MRI have been recently proposed for the diagnosis and staging of patients with breast cancer. FDG PET has proven to be effective in detecting distant metastases in patients with breast cancer [2]. However, the role of PET in assessing primary breast tumor is somewhat uncertain because of the variable sensitivity and specificity values that have been reported in the literature. This is mainly due to the lower metabolic activity of certain types of breast cancer compared with other malignancies, in particular using a single-time-point imaging technique [3]. Several studies have shown that the uptake of FDG continues to increase for several hours after its injection in different malignancies [4], and preliminary reports have shown the advantages of dual-time-point FDG PET in patients with primary breast tumors [5, 6]. In addition, integrated PET/CT machines that enable serial acquisition and subsequent display as a single fused image are commercially available and for breast imaging have shown some improvements over FDG PET alone [7]. MRI is increasingly being used to examine patients suspected of having breast cancer [8, 9]. The advantages of MRI for diagnostic evaluation of breast cancer include high soft-tissue contrast; multiplanar sectioning that permits the acquisition of contiguous thin sections, enabling a full 3D representation of one or both breasts; and the absence of ionizing radiation. The sensitivity of contrast-enhanced MRI for detecting breast cancer is generally high; and the specificity for lesion characterization that initially was only low to moderate has improved significantly in recent years [10–13].
This study was undertaken to assess the diagnostic value of dual-time-point FDG PET/CT performed with the patient in the prone position, in comparison with dynamic contrast-enhanced MRI in patients with a suspected breast malignancy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Mammography
Bilateral two-view mammography (Alpha RT, Instrumentarium) was performed at
our institution in craniocaudal and mediolateral oblique projections with spot
compression and with the acquisition of additional views when appropriate.
Mammographic findings were reported by an expert breast imaging radiologist
with 15 years of experience in breast imaging using the American College of
Radiology Breast Imaging Reporting and Data System (BI-RADS) categories
[14].
PET/CT Technique
All patients fasted for at least 6–8 hours before the PET examination
and had a blood glucose level < 140 mg/dL at the time of injection. FDG
(5.2 MBq/kg of body weight) was administered IV through an indwelling catheter
inserted into an antecubital vein contralateral to the site of the suspicious
lesion. Patients were kept quiet and comfortable for 60 minutes after the
injection. No muscle relaxants were administered. Immediately before the PET
acquisition, all patients were asked to empty their bladders because patients
were imaged without an indwelling urinary catheter. Images were obtained on a
single PET scanner (Discovery LS, GE Healthcare). The PET examination included
an initial whole-body acquisition with the patient in the supine position,
head to pelvis, and 4-MDCT with acquisition parameters of 140 kVp, 80 mAs, 0.8
second per rotation, 6:1 pitch, and slice thickness of 4.25 mm, performed
during normal breathing. This was followed by a PET acquisition of the same
axial range for 4 minutes per bed position. Immediately after the whole-body
scan, a second set of images of the breast in the prone position was acquired
(time point 1). The acquisition time for the emission scan was 5 minutes per
field of view, and both arms were positioned above the head to ensure free
hanging of the breasts, thereby avoiding compression and deformation. After 3
hours, a second PET acquisition was performed in the prone position (time
point 2). To minimize patient motion artifacts and to avoid variable breast
placement between examinations, a breast holder was used for both PET/CT
acquisitions. The CT data were used for attenuation correction, and images
were recon structed using a standard iterative algorithm. The duration of
dual-time-point PET/CT was well tolerated by all patients in the study.
The absorbed dose resulting from the IV administration of FDG was computed using dose coefficients provided by the International Commission on Radiological Protection (ICRP) in its Publication 80 for a variety of organs and tissues of the adult hermaphrodite medical internal radiation dose (MIRD) phantom [15]. To estimate radiation exposure of patients resulting from the acquisition of topograms and scans in CT, dose measurements were performed on an anthropomorphic whole-body RANDO phantom (Alderson Research Laboratories) using thermoluminescent dosimeters (TLDs). The method has been described in detail in a previous article [16].
PET/CT Data Analysis
Data were analyzed on an image fusion workstation (XELERIS, GE Healthcare)
in the coron al, sagittal, and axial planes as well as in 3D projections
(maximum intensity projection). Semiquantitative analysis was performed by one
nuclear medicine physician with 10 years of experience in nuclear medicine and
included assessing both sets of images at the same time. After image
reconstruction, regions of interest (ROIs) were carefully drawn around the
site of the PET lesion on the consequent four to six PET slices. Slice
thickness and slice interval were both 4 mm. ROIs were drawn on axial planes;
no 3D ROIs were used. The same ROI was used for both time point 1 and time
point 2. From these ROIs, the standard uptake value (SUV) was calculated
according to the formula (mean ROI activity in MBq/g / injected dose in MBq) /
body weight in g, where g is grams. The maximum SUV of FDG was measured from
the ROIs that were placed at the site of the lesion clearly visualized or
appear ing suggestive on the PET scan on the first time point
(SUVmax1) and on the second time point (SUVmax2). In
addition, the average percentage of change in SUVs (%SUVmax)
between time point 1 and time point 2 was calculated for PET/CT. All lesions
showing an SUVmax 2.5 on the first time point or an increase
in SUVmax values between time points 1 and 2 were considered
malignant; conversely, lesions showing an SUVmax < 2.5 on the
first time point or a decrease in SUVmax values between time points
1 and 2 were considered to be benign.
MRI Technique
MR examinations were performed using a 1.5-T whole-body MRI system
(Gyroscan Intera, Philips Healthcare) equipped with high-performance gradi
ents (master gradients), with a maximum slew rate of 150 (mT x
m–1) per millisecond and a maxi mum gradient strength of 30
mT/m. A dedicated four-element SENSE-compatible breast surface coil (MRI
Devices, In Vivo Research) was used with no breast fixation. The protocol con
sisted of a fast localizer (scout view) fast-field echo sequence with 18
sections (trans verse, coronal, and sagittal) using the following parameters:
TR/TE, 7.1/3.5; flip angle, 50°; field of view, 450 mm; and acquisition
time, 16.8 seconds. This sequence was used to prescribe sections of the
subsequent dynamic series to exactly cover the volume of fibroglandular tissue
of both breasts. The dynamic series consisted of an axial T1-weighted 3D
fast-field echo sequence (8.5/4.6; flip angle, 20°; and six dynamic
acquisitions), one obtained before and five obtained immediately after a bolus
injection of 0.1 mmol/kg of body weight of gadolinium dime glumine (Magnevist,
Schering) at an injection rate of 3 mL/s. Each dynamic volume consisted of 60
sections (2.5 mm thick) with an acquisition matrix of 272 x 242 and a
field of view of 360 mm (adjusted to the size of the breasts). With this
parameter setting, spatial resolution was 0.70 x 0.70 x 2.50 mm
and temporal resolution, 1 minute 1 second per each dynamic acquisition. To
suppress the signal of fat, image subtraction was performed offline after the
actual imaging session.
MRI Data Analysis
Images obtained were transferred to a dedicated workstation (Viewforum,
Philips Healthcare). Image interpretation was performed prospectively by a
radiologist who had 12 years of expertise in interpreting MR images of the
breast. The reader noted the number of lesions and assigned a BI-RADS category
for each lesion. Conventional images as well as mammograms, sonograms, and
reports were made available during the reading session just as they would be
in the clinical setting. The diagnostic criteria that were used to classify
lesions were based on lesion morphologic features (shape, margins, and
internal architecture) and lesion enhancement kinetics (enhancement rate in
the early contrast-enhanced phase and the signal intensity–time course
pattern in the intermediate and late contrast-enhanced phases). BI-RADS
category 1 was assigned if there was no enhancement at all; and category 2, if
the lesion shape and margins (oval mass with smooth borders) suggested a
benign lesion and if the internal architecture showed low-signal-intensity
internal dark septations (fibroadenoma) irrespective of enhancement rates or
time course kinetics. BI-RADS category 2 was also assigned if the internal
architecture was homogeneous or slightly heterogeneous and if the shape,
margins, and time course kinetics suggested a benign lesion (slow enhancement
rate and type 1 time course, with persistent enhancement and signal intensity
increasing steadily throughout the dynamic period. BI-RADS category 3 was
assigned if the shape and margins were not suspicious, enhancement rates were
fast, and a type 1 or 2 time course was observed; type 2 time course was
assigned when peak signal intensity was reached in the early contrast-enhanced
period and was followed by a plateau of signal intensity in the remaining
dynamic series. BI-RADS category 4 was assigned if, in the same setting, a
type 3 time course (washout) with peak signal intensity was reached in the
early phase and immediately followed by a loss of signal intensity in the
early contrast-enhanced period. BI-RADS category 4 was also assigned in
lesions with a suspicious irregular shape and margins status but
benign-appearing kinetics (slow enhancement rate and a type 1 time course).
BI-RADS category 5 was assigned if both the morphologic features (irregular
borders or spicules) and the time course kinetics indicated a malignant
lesion. BI-RADS category 5 was also assigned as soon as rim enhancement or
nonmasslike enhancement with asymmetric and segmental or ductal configuration
was noted, irrespective of shape, margins, enhancement rates, and time course
kinetics [17,
18]. The final diagnosis was
established by means of excisional or core biopsy in all patients.
Statistical Analysis
The findings of imaging studies were classified as true-positive,
true-negative, false-positive, or false-negative for malignancy on the basis
of histopathologic results. Sensitivity, specificity, accuracy, positive
predictive value, and negative predictive value were determined for both
imaging techniques using standard criteria
[19]. A p value of
less than 0.05 was considered statistically significant. Agreements among
dual-time-point PET/CT, single-time-point PET/CT, and MRI were quantified
using the kappa statistics. A kappa value of 0.20 or less indicated poor
agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate
agreement; 0.61–0.80, good agreement and 0.81–1.00, very good
agreement.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
Conversely, dual-time-point PET/CT showed an overall accuracy of 84% for
lesions with an SUVmax 2.5 or with a positive
%SUVmax, with sensitivity and specificity of 80% and 100%
versus 69% accuracy, 62% sensitivity (both, p < 0.001), and 100%
specificity (p = not significant) for single-time-point PET/CT
(Table 1). Overall accuracy,
sensitivity, and specificity of dual-time-point PET/CT for lesions > 10 mm
(n = 35) were 89%, 88%, and 100% versus 77%, 75%, and 100% for
single-time-point PET/CT (Table
2). For lesions < 10 mm (n = 20), dual-time-point
PET/CT showed an overall accuracy, sensitivity, and specificity of 75%, 62%,
and 100% versus 55%, 31%, and 100% of single-time-point PET/CT
(Table 3).
MRI showed two false-positives (both fibroadenomas with maximum diameters
of 9 and 11 mm) and one false-negative (lobular infiltrating carcinoma of 7
mm). Conversely, dual-time-point PET/CT showed nine false-negative (three
ductal carcinomas and two lobular carcinomas < 10 mm and three ductal
carcinomas and one lobular carcinoma > 10 mm). Single-time-point PET/CT
showed 17 false-negative (six ductal and three lobular carcinomas < 10 mm
and six ductal and two lobular carcinomas > 10 mm). On PET/CT, malignant
lesions showed an increase in FDG uptake between time points 1 and 2, with a
%SUVmax of 11 ± 24 (p < 0.04)
(Fig. 1). Conversely, benign
lesions showed either no change or a decrease in SUVmax between
time points 1 and 2, with a %SUVmax of –21 ± 7
(p < 0.001) (Fig.
2).
|
|
= 0.54) and fair agreement between MRI and single-time-point PET/CT
( = 0.34). Good agreement was reached between dual-time-point and
single-time-point PET/CT ( = 0.71). For lesions > 10 mm, fair
agreement was obtained between MRI and dual-time-point PET/CT ( = 0.34)
and between MRI and single-time-point PET/CT ( = 0.21); good agreement
was obtained between dual-time-point PET/CT and single-time-point PET/CT
( = 0.72). For lesions < 10 mm, good agreement was obtained between
MRI and dual-time-point PET/CT ( = 0.61), fair agreement between MRI
and single-time-point PET/CT ( = 0.33), and moderate agreement between
dual-time-point and single-time-point PET/CT ( = 0.58). Tables 1, 2, and 3 summarize the overall MRI, early PET, and late PET results in the patient population and according to lesion size. Figures 3A, 3B, 3C, 3D and 4A, 4B, 4C, 4D show examples of ductal infiltrating adenocarcinoma on the axial fused PET/CT image obtained in the prone position at time points 1 and 2 and the corresponding SUV values, and Figure 5A, 5B, 5C, 5D shows an example of lobular infiltrating adenocarcinoma. These figures also show subtracted contrast-enhanced T1-weighted axial MR images and dynamic contrast-enhanced MR signal intensity curves.
|
|
|
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In addition, several studies have recently shown that uptake of FDG continues to increase for several hours after injection in different malignancies [4], and investigators from the University of Pennsylvania have reported initial results of dual-time-point imaging in animal models and patient studies [23, 24]. In particular, Hustinx et al. [24] performed dual-time-point scanning of 21 patients with 18 malignant head and neck tumors and nine inflammatory or infectious lesions. They noted that tumors had an average SUV increase of 12% between the first and second scans, whereas inflammatory lesions and structures with physiologic uptake of FDG (tongue, larynx) showed essentially stable uptake or a slight decline over time. Another important finding was that the SUV changes in tumors were greater when more than 30 minutes had elapsed between the first and second emission scans. Most breast malignancies also show a gradual increase in SUV values over time after FDG injection, and recent preliminary reports support this hypothesis [5, 6].
In agreement with these studies, in our study malignant lesions showed a
significant increase in FDG over time compared with benign lesions. These
changes in SUVs might be considered a reliable parameter that can be used to
differentiate benign from malignant lesions of the breast on PET/CT. In
particular, for lesions with an SUVmax 2.5 or with a positive
%SUVmax, dual-time-point PET/CT with acquisition in the
prone position showed a significantly higher accuracy (84%) than
single-time-point PET/CT (69%). An SUV threshold value of 2.5 has been cited
in several articles as the optimal threshold to differentiate benign from
malignant tumors. Most inflammatory lesions fall below this threshold, whereas
most malignant lesions have SUV values greater than 2.5
[25,
26]. A previous study
[5] showed that inflammatory
lesions are indistinguishable from normal breast tissue by dual-time-point
analysis because both show either no change or negative dual-time-point
changes in most SUVs. Therefore, the change in SUVs over time estimated with
the dual-time-point technique is helpful in differentiating breast cancer from
inflammatory disease. In addition, dual-time-point imaging can improve the
sensitivity of PET, especially in mammographically dense breasts, which
normally show increased FDG uptake
[27]. Thus, the application of
dual-time-point imaging of the breast can improve overall accuracy of the
technique by correctly identifying benign normal tissue and inflammatory
lesions that at times may mimic cancer.
The ability of PET to detect breast cancer also depends greatly on tumor size; in our study, overall sensitivity of dual-time-point PET/CT for lesions > 10 mm (n = 35) was 88%, versus 75% for single-time-point PET/CT. Furthermore, for lesions < 10 mm (n = 20), dual-time-point PET/CT showed an overall sensitivity of 62% versus 31% for single-time-point PET/CT. These results are in agreement with previous data by Avril et al. [28], who showed that only 30 (68.2%) of 44 breast carcinomas were identified on FDG PET at stage pT1 (< 2 cm) compared with 57 (91.9%) of 62 at stage pT2 (> 2–5 cm). Sensitivity for tumors smaller than 1 cm (pT1a and pT1b) was only 25%, compared with 84.4% for tumors between 1 and 2 cm in diameter (pT1c). Although the dual-time-point system PET/CT partially solves the problems related to lesion localization, the spatial resolution of the technique is still limited, and the high cost discourages the routine clinical application of FDG PET/CT in the screening and diagnosis of primary breast tumors, especially for lesions < 1 cm.
A dedicated PET breast scanner has recently been developed to improve both resolution and contrast of breast lesions and thereby improve small lesion detectability while reducing the cost of imaging [29]. By optimizing scanner geometry for breast imaging, a dedicated breast PET scanner can achieve much higher resolution than a whole-body scanner and still maintain higher sensitivity and a marked reduction in scatter. The final output resolution of commercially available positron emission mammography (PEM) is approximately 1.5 mm in-plane. With this high resolution, PEM can detect smaller objects than would be detectable using a whole-body scanner [30]. A recent multicenter study showed that high-resolution FDG PEM detects in situ components of cancers better than any other technique; this fact has been documented in retrospective surgical studies [31]. Therefore, advances in technology such as the development of dedicated breast imaging devices may further improve the future detection of primary tumors with PET.
Contrast-enhanced MRI has been shown to be highly effective for the detection and characterization of suspected breast disease, with reported sensitivities of up to 90% [8, 9]; in particular, the rapid improvements in MRI hardware and software design have made this technique a viable complementary examination to conventional mammography and sonography. Although mammography is the study of choice for screening, breast MRI is preferred for the local staging of breast cancer. MRI allows the most accurate delineation of the size and local extent of cancer, including the depiction of multifocal, multicentric, or contralateral disease, and offers the highest sensitivity for showing intraductal extension around invasive cancers. Furthermore, because of its high negative predictive value, MRI can be used to confidently exclude the presence of breast cancer and thus avoid unnecessary surgery. For all these reasons, MRI should be considered an integral part of the conventional workup of patients with breast cancer [32].
Unfortunately, MRI has yet to become a widely accepted procedure for breast imaging because of its diverse specificity data [33, 34]. A recent study comparing MRI of the breast with FDG PET found a comparable diagnostic accuracy (88% vs 84%) for both imaging techniques in 32 patients [35]. The sensitivity of FDG PET was 79%, whereas MRI detected all primary breast carcinomas; however, the specificity of FDG PET was higher (94% vs 72%). In agreement with these data, in our study breast MRI showed a trend toward higher accuracy and sensitivity and limited specificity compared with PET/CT; specificity was reduced because of the number of benign lesions that exhibited strong and early enhancement after IV administration of gadopentetate dimeglumine. This enhancement pattern is reported to be common in rapidly growing fibroadenomas because of their hyaline content and because of the increased vascularity compared with that of breast parenchyma, as well as in patients with highly proliferative dysplasia and in patients with abscesses and fibrocystic changes [11, 12, 36]. The higher sensitivity of MRI over PET/CT, as observed in our study, is probably due to its better spatial resolution that allows better detection of lesions < 10 mm.
Potential limitations of dual-time-point PET/CT are its low availability and the radiation exposure of this technique compared with other breast imaging techniques such as MRI. However, in our study, dual-time-point PET/CT showed an overall effective dose of 26.8 mSv, which is slightly higher than that observed with single-time-point PET/CT (21.4 mSv) but not significantly different from that (25 mSv) obtained with diagnostic PET/CT at 120 kVp and 200 mA [37]. Furthermore, MRI and FDG PET/CT remain limited diagnostic techniques in detecting in situ or very small (< 3 mm) breast carcinomas because the neoangiogenesis induced by these small or intraductal tumors is too faint to be detected on contrast-enhanced MRI or PET/CT. Therefore, it is important to emphasize that the final diagnosis should always be based on a combination of the gadolinium enhancement or FDG uptake pattern, morphology and architecture of the lesions, and clinical and mammographic or sonographic information. All cases with discrepancies between morphologic, MR signal intensity, and FDG uptake findings require further investigation with excisional biopsy.
Finally, to our knowledge, this is the first study in the English-language literature to directly compare FDG PET/CT performed in prone position and breast MRI in patients with suspected breast cancer.
In conclusion, dual-time-point imaging with acquisition in the prone
position improves PET/CT accuracy in patients with suspected breast malignancy
over single-time-point PET/CT. In particular, the uptake of FDG increases over
time in malignant breast tumors and decreases over time in benign lesions.
Dual-time-point PET/CT performed in the prone position should be preferred to
single-time-point PET/CT and is recommended for imaging patients with
suspected breast malignancy. However, the limited sensitivity of FDG PET/CT,
especially for lesions 
10 mm, as observed in our study, suggests that
PET/CT cannot be used as a routine imaging procedure for patients with
suspected breast carcinoma and cannot significantly reduce the necessity of
invasive procedures in patients suspected of having primary breast cancer. MRI
shows higher sensitivity and lower specificity than PET/CT for disclosing
breast malignancy and should be preferred for the detection and
characterization of lesions 10 mm.
The development of new technology combining the higher sensitivity of MRI and the higher specificity of PET may further improve the diagnostic accuracy of noninvasive imaging techniques in patients with suspected breast carcinoma.
Acknowledgments
We thank Graciana Diez-Roux for critically reviewing this manuscript.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Q. Maguire Jr., M. E. Noz, and F. Ponzo Prone Breast Dual-Time-Point PET/CT Compared With MRI for Determining Breast Cancer Am. J. Roentgenol., July 1, 2009; 193(1): W77 - W77. [Full Text] [PDF]
|
|
|
|
|
|
M. Imbriaco, F. Soscia, and G. Avitabile Reply Am. J. Roentgenol., July 1, 2009; 193(1): W78 - W78. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
