| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiological Sciences, University of Rome "La
Sapienza," Viale le Regina Elena 324, 00161 Rome, Italy.
2 Worldwide Medical & Regulatory Affairs, Bracco Imaging SpA, Milano,
Italy.
3 Worldwide Medical & Regulatory Affairs, Bracco Diagnostics, Inc.,
Princeton, NJ.
Received December 11, 2007;
accepted after revision May 16, 2008.
Address correspondence to F. Pediconi
(federica.pediconi{at}uniroma1.it).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty-seven women (mean age ± SD, 50.8 ± 12.9 years) with breast lesions classified as BI-RADS category 3, 4, or 5 for suspicion of malignancy underwent two identical MR examinations at 1.5 T separated by 48–72 hours. T1-weighted gradient-echo images were acquired before contrast administration and at 2-minute intervals after the randomized injection of gadopentetate dimeglumine or gadobenate dimeglumine at 2 mL/s. Two blinded readers evaluated randomized image sets for lesion detection and differentiation as benign or malignant compared with histology. The McNemar exact test and the generalized estimating equation (GEE) were used to compare lesion detection rates and diagnostic performance in terms of sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV).
RESULTS. Histopathology data were available for 78 lesions. Significantly more lesions overall (75/78 [96%] vs 62/78 [79%], respectively; p = 0.0002) and significantly more malignant lesions (49/50 [98%] vs 38/50 [76%]; p = 0.0009) were detected with gadobenate dimeglumine than gadopentetate dimeglumine. All detected malignant lesions were correctly diagnosed with both agents. More detected benign lesions were correctly diagnosed with gadobenate dimeglumine than with gadopentetate dimeglumine (20/26 [77%] vs 17/24 [71%], respectively). Differentiation of lesions was significantly (p = 0.0001) better with gadobenate dimeglumine. Significantly better diagnostic performance was noted with gadobenate dimeglumine than with gadopentetate dimeglumine, respectively, for sensitivity (98.0% vs 76.0%; p = 0.0064), accuracy (88.5% vs 69.2%; p = 0.0004), PPV (86.0% vs 76.0%; p = 0.0321), and NPV (95.2% vs 57.1%; p = 0.0003).
CONCLUSION. Lesion detection and malignant–benign differentiation is significantly better with 0.1 mmol/kg gadobenate dimeglumine than 0.1 mmol/kg gadopentetate dimeglumine.
Keywords: breast cancer • breast MRI • breast neoplasms • contrast media • gadobenate dimeglumine • MR mammography
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
4.3–5.0 L x mmol–1 x s–1
at 1.5 T), and thus similar lesion signal intensity enhancement and dynamic
contrast behavior are seen when these agents are administered at an equivalent
dose [19,
20]. Gadobenate dimeglumine (MultiHance, Bracco Imaging) is a contrast agent with increased R1 relaxivity relative to conventional gadolinium agents due to a weak, transient interaction of the contrast-effective chelate of gadobenate dimeglumine (Gd-BOPTA) with serum albumin [21, 22]. Compared with gadopentetate dimeglumine, the R1 relaxivity of gadobenate dimeglumine in blood is roughly twofold higher at all commercially available magnetic field strengths, ranging from 10.9 versus 4.7 L x mmol–1 x s–1 at 0.2 T to 6.3 versus 3.3 L x mmol–1 x s–1 at 3 T [23]. Numerous studies in a variety of clinical applications have shown that the increased R1 relaxivity of gadobenate dimeglumine translates into greater signal intensity enhancement and thus improved image quality and diagnostic performance compared with comparator contrast agent at an equivalent or higher dose [24–28]. The possibility to achieve greater contrast enhancement using a standard dose of 0.1 mmol/kg of body weight compared with that achievable with an equivalent dose of a conventional gadolinium agent may have particular value for imaging applications in the breast where lesions often are small, poorly enhancing, or otherwise inconspicuous against the surrounding normal breast parenchyma [29, 30].
The results of a previous intraindividual crossover study of 25 consecutive women revealed that 0.1 mmol/kg of body weight of gadobenate dimeglumine was superior to an equivalent dose of gadopentetate dimeglumine for both detection and characterization of breast lesions [31]. However, a possible limitation of that earlier study is that only 46 lesions were evaluated, of which only eight were benign. The present intraindividual crossover study in a larger patient population and involving a larger overall number of histopathologically proven lesions and a greater proportion of benign lesions was aimed at further confirming the value of gadobenate dimeglumine relative to gadopentetate dimeglumine for the detection and accurate differentiation of benign from malignant breast lesions.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Enrolled patients each underwent two identical breast MRI examinations, one with gadobenate dimeglumine as the MR contrast agent and the other with gadopentetate dimeglumine. The order in which the agents were administered was fully randomized: 23 patients (mean age ± SD, 51.3 ± 11.8 years; range, 32–74 years) received gadobenate dimeglumine for the first examination and gadopentetate dimeglumine for the second (group A), and 24 patients (mean age, 50.4 ± 14.1 years; range, 30–75 years) received gadopentetate dimeglumine for the first examination and gadobenate dimeglumine for the second (group B).
Surgery (lumpectomy, quadrantectomy, or mastectomy), excisional biopsy, or core biopsy with histologic evaluation of pathologic specimens was performed in all patients from 24 hours to 1 month after completion of the second MRI examination. Core biopsy and needle localization for lumpectomy and excisional biopsy in all cases were performed under imaging guidance (mammography or sonography) to better visualize the lesion. For the purposes of the study, a lesion was defined according to the BI-RADS lexicon as a mass, an area of non–masslike enhancement, or a focus [2, 12]. All lesions considered category B3 (borderline for malignancy) according to the BI-RADS classification were referred for excisional biopsy after core biopsy; core biopsy was accepted for only the lesions that were classified as truly benign. Although there is some controversy among the radiologic community regarding the classification of lobular carcinoma in situ (LCIS) [32–39], at our institution these lesions are considered high risk and are always resected after positive diagnosis on excisional biopsy. Accord ingly, LCIS was considered a malignant lesion for the purposes of this study. This approach has been adopted previously [31].
Follow-up of all patients was performed using routine mammography, sonography, and MRI after an interval of 12–24 months, depending on the age of the patient.
MRI
Breast MRI was performed on a 1.5-T magnet (Vision Plus, Siemens Medical
Solutions) using a bilateral breast surface coil with the patient in the prone
position. An axial 3D dynamic T1-weighted gradient-echo sequence was used with
images acquired before contrast agent administration (i.e., unenhanced images)
and 0, 2, 4, 6, and 8 minutes after administration of the contrast agent
(i.e., contrast-enhanced images). The imaging para meters were identical for
both MRI examinations in each patient, with a TR/TE of 8.1/4, flip angle of
30°, 1 excitation, a rectangular field of view of 
36 cm, and matrix
of 396 x 512. The slice thickness for each patient was 2 mm and no
interslice gap was used. The total acquisition time was approximately 90
seconds for each sequence acquisition. No fat-suppression, parallel imaging,
or partial-Fourier sequences were used. Image subtraction (postcontrast
– precontrast on a pixel-by-pixel basis) was performed to eliminate the
signal of fat. The total voxel size was 1 x 0.6 x 1.6 mm.
For each examination, the contrast agent was administered at an identical flow rate of 2 mL/s (bolus) by means of a power injector (Spectris, Medrad) through an antecubital vein. Each con trast agent was administered at a final dose of 0.1 mmol/kg of body weight and was followed by a 20-mL saline flush. Data acquisition started after con trast agent injection, at the end of the saline flush.
The interval between the two MRI examinations was in all cases greater than 48 hours but less than 72 hours. A minimum interval of 48 hours between the two examinations was considered necessary to ensure complete elimination of the first contrast agent before administration of the second [25, 26], and a maximum interval of 72 hours was considered appropriate to ensure full comparability of imaging findings from the two examinations.
Image Assessment
All images were evaluated by two observers (6 and 7 years' experience of
breast MRI interpretation) in consensus. The observers were unaware of all
patient data (identity, medical history, clinical profile, results of other
imaging procedures) and were fully blinded to the identity of the contrast
agent administered. To minimize recall bias, images from the second MRI
examination were assessed 2 weeks after those obtained from the first
examination. Each evaluated image set from each examination comprised
unenhanced, contrast-enhanced, and subtracted (postcontrast –
precontrast) images and additional maximum-intensity-projection (MIP)
reconstructions of subtracted images at the first postcontrast time point.
Assessment of images in each reading session was performed with the image
sets presented in fully randomized order and was conducted as performed
routinely in daily clinical practice. Image sets were evaluated first for
technical quality using a subjective adequate or not adequate criterion; if
the observers considered an image set to be inadequate because of motion
artifacts, no further assessments were performed. The position and size of the
lesions detected on each technically adequate image set were thereafter
indicated on breast maps for later use in lesion matching. Lesion size was
characterized using the following 5-point scale: 1 = 5 mm, 2 = 6–10
mm, 3 = 11–20 mm, 4 = 21–50 mm, and 5 = > 51 mm. Lesion
matching was performed by a third observer (4 years' experience) who played no
part in the initial assessment. The purpose was to match any lesions detected
during the first MRI examination with lesions detected during the second
examination. The information indicated for each lesion included location
(breast quadrant) and size.
The imaging findings on each image set from each breast MRI examination were subsequently evaluated against histopathology findings in terms of overall detection of lesions and their accurate differentiation as malignant or benign. Detected lesions were classified as malignant or benign on the basis of lesion morphology using the BI-RADS 6-point scale for suspicion of malignancy [12] and by assessing postcontrast signal intensity–time curves obtained at up to three regions of interest placed on enhancing regions. The signal intensity–time curves were classified according to shape, from type I to type III as described elsewhere [40–42].
Statistical Analysis
The lesion detection rate was defined as the number of breast lesions
detected on MRI divided by the number of histologically confirmed breast
lesions. Lesion detection rates after gadobenate dimeglumine and gadopentetate
dimeglumine were compared using the McNemar exact test. The same test was used
to compare the detection rates for malignant and benign lesions separately.
For the purposes of the study, a true-positive lesion was a lesion that was
characterized as malignant on MRI and was confirmed as malignant at histology.
Likewise, a true-negative lesion was a lesion that was characterized as benign
(non-malignant) on MRI and was confirmed as benign at histology. Conversely, a
lesion was considered a false-positive if it was detected and characterized as
malignant at MRI but either was not detected or was detected and confirmed as
benign at histology. Finally, a false-negative lesion was a lesion that was
not detected or was detected and characterized as benign at MRI and was
detected and confirmed as malignant at histology.
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Lesion Detection
Overall, significantly (p = 0.0002, McNemar exact test) more
lesions were detected on breast MRI after gadobenate dimeglumine
administration (75/78 [96.2%]) than on MRI after gadopentetate dimeglumine
administration (62/78 [79.5%]). The principal benefit of gadobenate
dimeglumine was for the detection of the 50 malignant lesions: 49 (98%) were
detected after gadobenate dimeglumine compared with only 38 (76%) seen after
gadopentetate dimeglumine (p = 0.0009)
(Fig. 1). The malignant lesion
not detected after gadobenate dimeglumine was a small ( 5 mm) DCIS in a
patient with two DCIS lesions confirmed after lumpectomy. Conversely, the 12
malignant lesions not detected after gadopentetate dimeglumine comprised the
same DCIS lesion plus eight small invasive ductal carcinoma lesions (four
lesions of 5 mm, four lesions of 6–10 mm) in six patients (Fig.
2A,
2B,
2C,
2D,
2E,
2F) (one lesion in each of four
patients; two lesions in each of two patients), two small invasive lobular
carcinoma lesions (6–10 mm) in two patients (Fig.
3A,
3B,
3C,
3D,
3E), and one LCIS lesion
(6–10 mm) in one patient. In each case except the LCIS, the malignant
lesions missed with gadopentetate dimeglumine occurred in patients with
histopathologically confirmed multifocal or multicentric disease. In four
patients, two lesions were detected with gadobenate dimeglumine compared with
just one lesion (three patients) or no lesions (one patient) with
gadopentetate dimeglumine; in four other patients, three lesions (two
patients) or four lesions (two patients) were detected with gadobenate
dimeglumine compared with two lesions (three patients) or three lesions (one
patient) with gadopentetate dimeglumine.
|
|
|
|
|
|
|
|
|
|
|
5 mm
with gadopentetate dimeglumine was reported as being 6–10 mm with
gadobenate dimeglumine.
Six lesions considered present on gadobenate dimeglumine–enhanced MR images and five lesions considered present on gadopentetate dimeglumine–enhanced MR images were not present at pathology (i.e., lesions were false-positive for detection). The six lesions after gadobenate dimeglumine were considered malignant in four cases and benign in two cases, whereas the five lesions after gadopentetate dimeglumine were considered malignant in three cases and benign in two cases.
Lesion Characterization
The 49 malignant lesions detected after gadobenate dimeglumine
administration and the 38 malignant lesions detected after gadopentetate
dimeglumine administration were in all cases correctly classified as malignant
before histopathologic assessment. Conversely, only 20 (76.9%) of the 26
benign lesions detected after gadobenate dimeglumine and only 17 (70.8%) of
the 24 benign lesions detected after gadopentetate dimeglumine were correctly
classified as benign on MRI before histopathology. The misdiagnosed benign
lesions included two papillomas in two patients and one adenosis, one atypical
hyperplasia, and one fibrocystic change in one patient each that were
misdiagnosed as malignant after both contrast agents. Differences between the
two contrast agents concerned one papilloma that was misdiagnosed as malignant
after gadobenate dimeglumine but was not detected at all after gadopentetate
dimeglumine and two additional fibrocystic changes in one patient that were
misdiagnosed as malignant only after gadopentetate dimeglumine.
Based on evaluation of all 78 histopathologically confirmed lesions, gadobenate dimeglumine was significantly (p = 0.0001, McNemar exact test) superior to gadopentetate dimeglumine for the correct differentiation of breast lesions as malignant or benign. The diagnostic performance of breast MRI for the correct characterization of lesions is summarized in Table 1 for both contrast agents. Although no significant difference was noted for specificity (p = 0.1277), reflecting the comparatively small number of evaluated lesions, significant superiority for gadobenate dimeglumine compared with gadopentetate dimeglumine was noted on GEE analysis for sensitivity (98.0% vs 76.0%; p = 0.0064), accuracy (88.5% vs 69.2%; p = 0.0004), PPV (86.0% vs 76.0%; p = 0.0321), and NPV (95.2% vs 57.1%; p = 0.0003).
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The results of our study suggest that careful and appropriate selection of the MR contrast agent can also significantly benefit the diagnostic performance of breast MRI. Specifically, the results of the present study confirm the results of a previous smaller-scale study [31] in showing that significantly improved breast lesion detection can be achieved with gadobenate dimeglumine compared with gadopentetate dimeglumine when these agents are used at a standard dose of 0.1 mmol/kg of body weight.
Of importance is that the greatest benefit seen was for the detection of malignant lesions; all but one histopathologically proven malignant lesions (49/50 [98%]) were detected with gadobenate dimeglumine, whereas only 38 (76%) of 50 malignant lesions were detected with gadopentetate dimeglumine. Similar results were noted previously (45/46 [98%] malignant lesions detected with gadobenate dimeglumine vs 36/46 [78%] malignant lesions with gadopentetate dimeglumine; p = 0.003) and were attributed to the higher R1 relaxivity of gadobenate dimeglumine in blood and hence to a greater lesion signal intensity enhancement at an equivalent dose [31]. Similar conclusions regarding the greater signal intensity enhancement achievable with gadobenate dimeglumine have been drawn elsewhere for intraindividual crossover studies with gadopentetate dimeglumine in the CNS [27, 28], vasculature [24–26], liver [46], and myocardium [47].
The detection of histopathologically proven benign lesions was also better with gadobenate dimeglumine (26/28 [92.8%]) than with gadopentetate dimeglumine (24/28 [85.7%]). However, whereas all malignant lesions were correctly characterized as such on the basis of morphologic and contrast enhancement characteristics, only 20 (76.9%) of 26 and 17 (70.8%) of 24 proven benign lesions were correctly characterized as such after gadobenate dimeglumine and gadopentetate dimeglumine, respectively. Of interest is that none of the benign lesions misdiagnosed as malignant on breast MRI with either contrast agent was a fibroadenoma, which contrasts with previous findings both with gadobenate dimeglumine [31, 48] and other MR contrast agents [49–51].
The benign lesions that were misdiagnosed as malignant comprised mainly papillomas along with solitary cases of adenosis and atypical hyperplasia and fibrocystic changes. Although the misdiagnosis of these lesions was a somewhat surprising finding, the fact that five of these lesions were misdiagnosed equally after both contrast agents suggests that the reasons for misdiagnosis were primarily related to the lesion rather than the observer. In this regard, most of the misdiagnosed lesions were small—one of the misdiagnosed papillomas after gadobenate dimeglumine was not detected at all after gadopentetate dimeglumine—with poorly defined morphologic features and ambiguous dynamic enhancement.
In the case of papillomas, these lesions, like fibroadenomas, frequently mimic invasive breast cancer [51, 52], which may partly explain their misdiagnosis in this study. Moreover, it should be borne in mind that intraductal papillomas are often considered harbingers of risk, particularly if multiple and peripheral [53–55], and are invariably referred for regular follow-up if not resected immediately because of the possibility of malignant degeneration. In this setting, their diagnosis as malignant should not be considered a negative finding because all lesions diagnosed as malignant or high risk (i.e., including LCIS and papilloma) should be referred for presurgical biopsy and histologic evaluation. As noted elsewhere [56], from a clinical perspective the additional true-positive malignant lesions detected should certainly outweigh the occasional misdiagnosis of true-negative lesions, particularly if patients are considered at high risk for breast cancer.
Of note is that LCIS was considered a malignant lesion for the purposes of this study. Although there is currently a great deal of controversy among radiologists regarding the classification of LCIS [32–38], the widely accepted viewpoints are for LCIS diagnosed at surgical biopsy to be considered a lesion that does not need treatment but that indicates an increased risk of breast cancer occurring at any site in either breast [32], whereas LCIS diagnosed at needle biopsy is to be considered a high-risk lesion in need of excision to exclude associated cancer missed at the time of biopsy [33–35]. However, it has been reported that the risk of developing breast cancer is 8–11 times higher among women with LCIS compared with women without LCIS [39] and that 10–20% of patients with LCIS develop breast cancer within 15–25 years of the initial diagnosis [36]. Because of the lack of consensus agreement about the appropriate treatment option for patients with diagnosed LCIS [32–39], at our institution we routinely consider LCIS a malignant lesion for subsequent patient management decisions and treat it in a manner similar to the way in which DCIS is treated (i.e., clear margins at excision and use of radiation therapy in cases of lumpectomy).
In terms of overall diagnostic performance (i.e., lesion detection and correct differentiation as benign or malignant), significant superiority was noted for gadobenate dimeglumine compared with gadopentetate dimeglumine for sensitivity (98.0% vs 76.0%; p = 0.0064), accuracy (88.5% vs 69.2%; p = 0.0004), PPV (86.0% vs 76.0%; p = 0.0321), and NPV (95.2% vs 57.1%; p = 0.0003). Conversely, the difference in specificity (71.4% vs 57.1%) was not significant (p = 0.1277). Although specificity values of 100% have been reported previously for both gadobenate dimeglumine and gadopentetate dimeglumine, these values were considered artificially high because of the small number of benign lesions (n = 8) in the evaluated lesion population [31]. In this study, 28 benign lesions were evaluated, which was insufficient for detection of a significant benefit for gadobenate dimeglumine but was sufficiently large to provide a more clinically realistic assessment of the comparative specificity. Concerning the significant differences in PPV and NPV for the two contrast agents, this difference may be of particular relevance if breast MRI is ever to be accepted as a screening procedure for the general population and should certainly be of current interest for breast screening in women considered at high risk of developing breast cancer.
In showing the significant superiority of gadobenate dimeglumine compared with gadopentetate dimeglumine for breast MRI, this study confirms the conclusions of previous studies [31, 48, 57]. Nevertheless, the fact that only 47 patients were evaluated is still a potential limitation of the study. Future work should certainly be performed in a much larger, multicenter patient population to corroborate these findings.
Acknowledgments
We thank Ningyan Shen, Usha Halemane, and Riccardo Spezia for assistance
with the statistical analyses performed in this study.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Romano Radiology in Italy: What Is Happening? Am. J. Roentgenol., October 1, 2009; 193(4): W273 - W277. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
