| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Johannes Gutenberg University of Mainz,
Langenbeckstr. 1, Mainz D-55131, Germany.
2 Department of Pathology, Johannes Gutenberg University of Mainz, Mainz
D-55131, Germany.
Received August 26, 2002;
accepted after revision March 3, 2003.
Address correspondence to A. Teifke.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. MRI of the breast (1.0-T scanner, dynamic gadolinium-enhanced T1-weighted three-dimensional gradient-echo sequence, double breast coil) was performed on 1273 women for different indications. Enhancing incidental lesions were defined as enhancing lesions on MRIs that were not expected from findings on the previous conventional imaging. They were classified in five assessment categories using a scoring system based on morphologic and kinetic enhancement characteristics. Detection of enhancing incidental lesions resulted in a review of mammograms and sonograms with the aim of also localizing these lesions on conventional imaging. The lesions were either biopsied or followed up.
RESULTS. Twenty-five percent (274/1086) of all enhancing lesions detected in the study population were enhancing incidental lesions. Enhancing incidental lesions were found in 16% (210/1273) of all study patients. Forty-one percent (113/274) of the enhancing incidental lesions were histologically confirmed. Forty-eight percent (54/113) of the biopsied and 20% (54/274) of all enhancing incidental lesions were malignant. Eleven percent (54/508) of all malignant lesions occurring in the 1273 women were detected solely through additional MRI. Fifty-seven percent (31/54) of these MRI-detected malignant lesions could be identified on a reevaluation of sonograms and mammograms.
CONCLUSION. Detection of enhancing incidental lesions should lead to a thorough reevaluation of mammograms and sonograms. If not reidentified, suspicious enhancing incidental lesions should be biopsied, and enhancing incidental lesions that are probably benign should be carefully followed up. Indeterminate enhancing incidental lesions should be histologically examined by minimally invasive techniques or, if they are small, followed up by another MRI 6 months later.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
The second most common indication was the search for additional tumor foci (21% of examinations) when a high degree of suspicion for malignancy already existed on the basis of conventional breast imaging. Indeed, malignancy of the primary lesion was confirmed in all but one of these patients. Most of the remaining patients had breasts that were difficult to image on mammography and sonography because of severe scars or augmentation after breast cancer surgery (13% of examinations). Patients with suspicious microcalcifications (9% of examinations) were examined before biopsy for purely academic reasons to evaluate the enhancement characteristics of ductal carcinoma in situ (DCIS). Another 9% of the patients had dense breast tissue and a history hinting at an increased risk (e.g., familial breast cancer or lobular carcinoma in situ in previous biopsy).
The entire study protocol was approved by our regional review board. All patients gave informed consent to participate in this study.
In premenopausal patients (n = 509), the examinations were performed between days 5 and 15 of the menstrual cycle. In patients receiving hormone replacement therapy (n = 215), the medication was not interrupted before the examinations.
After MRI was performed, tissue sampling was recommended in all patients with suspicious lesions on mammography, sonography, or MRI. Equivocal lesions revealed on conventional imaging and negative results on MRI were normally not biopsied but were followed up on mammography or sonography at 6, 12, 24, and 36 months (e.g., patients with masses that proved to be cysts on MRI, asymmetric densities, or well-circumscribed masses on conventional imaging without any enhancement). All suspicious microcalcifications were histologically examined even in the case of negative MRI results because of the well-known limited sensitivity of MRI in cases of DCIS. Eventually, 831 lesions were histologically examined in 621 of the 1273 patients in this study. Of these 831 lesions, 508 (61%) turned out to be malignant.
Conventional Imaging
In all patients, high-quality mammograms (obtained with different equipment
and types of films because patients were scanned at different facilities) were
available. Before these patients underwent MRI, these mammograms had been
reviewed and a clinical examination and sonography (Sonolayer SSA-250A with
SMA-736SA annular array probe, Toshiba Medical Systems, Tokyo, Japan) of both
entire breasts had been performed in all patients by one of three radiologists
specialized in breast imaging, including MRI and nonsurgical breast biopsies.
In our institution, sonography is a standard procedure in all patients with
abnormal findings on mammography or with breast tissue that is almost entirely
fat [10]. Additional spot
compression and magnification mammograms (Senographe DMR, General Electric
Medical Systems, Milwaukee, WI) were routinely obtained if considered useful
in the diagnostic workup. Mammographic evaluation criteria were based on
BI-RADS. Sonographic criteria for malignancy were spiculated or ill-defined
margins, irregular shape, posterior acoustic shadowing, architectural
distortion, and solid intracystic or intraductal lesions.
MRI
MRI was performed with a 1.0-T imager (Magnetom Impact Expert 42SP/AS,
Siemens Medical Systems, Erlangen, Germany) and the manufacturer's double
breast coil with the patient in the prone position. At first, T2-weighted
coronal turbo spin-echo sequences (TR/TE, 5432/90; field of view, 175 x
350; resolution, 1.39 x 1.38 x 3 mm) were obtained. These were
followed by a dynamic contrast-enhanced coronal T1-weighted three-dimensional
gradient-echo sequence (15/7; flip angle, 30°; field of view, 350 x
175 x 119; resolution, 1.82 x 1.37 x 1.86 mm; sections, 64)
with an acquisition time of 93 sec. One measurement was acquired before, and
five consecutive measurements were acquired after a manual bolus injection of
gadopentetate dimeglumine (Magnevist, Schering, Berlin, Germany) in a dose of
0.1 mmol per kilogram of body weight. The system software was used to
calculate subtraction images from the first and the third contrast-enhanced
study and the unenhanced dynamic study. For enhancing lesions, intensity-time
curves were calculated in regions of interest with a workstation (SPARC
Station, Sun Microsystems, Mountain View, CA) and a software package (MR
Vision, Menlo Park, CA). The regions of interest (2-9 pixels) were placed in
the tumor area with the highest signal-intensity enhancement. The MRI
examinations were evaluated prospectively by two of the three attending
radiologists in a consensus reviewing.
Absence of contrast media uptake or a bilateral slow and diffuse enhancement was considered negative for cancer. A unilateral diffuse or segmental enhancement was considered suspicious. A special scoring system was applied for the grading of circumscribed enhancement foci (Table 2). This system is a modification of the classification model proposed by Fischer et al. [3]. It was adapted to our MRI parameters and optimized on a data set of 150 histologically diagnosed enhancing lesions. On the basis of this scoring system, each lesion was classified into the following categories: 2, benign finding; 3, probably benign finding; 4a, indeterminate finding; 4b, probably malignant; or 5, highly suggestive of malignancy. A bilaterally diffuse, patchy, or dotlike enhancement pattern was not considered abnormal. Mammographic and sonographic documentation were available during the evaluation. An enhancing lesion without correlate on the previous mammogram or sonogram was considered an enhancing incidental lesion.
|
Clinical Management
The detection of enhancing incidental lesions resulted in a thorough
retrospective analysis of mammograms and repeated sonograms with the aim of
also identifying and assessing the lesions by conventional technology. For
this purpose, multiplanar reconstructions of the MRIs and, if useful,
additional mammograms were obtained. The clinical procedure was decided on the
basis of the findings in this review, the score of the enhancing incidental
lesion, the general management concept, and the preference of the patient.
Histologic examination was recommended for all suspicious lesions. Likewise,
histologic evaluation was performed for several benign-appearing lesions when
this was desired by the patient or when a biopsy could be easily performed as
a core needle biopsy or during an independent surgical procedure. Except when
mastectomy was planned beforehand, all lesions that were solely identified on
MRI were preoperatively localized under MRI (Biopsy Mamma Coil, Noras,
Würzburg, Germany) or CT guidance (Figs.
1A,
1B,
1C, and
1D). Lesions that could be
correlated on reevaluation of conventional imaging with high certainty were
localized with mammographic or sonographic guidance with a repositionable
MRI-compatible needle. Afterwards MRI was performed to ensure that the lesions
were correctly identified. Histologically benign results in the case of
suspicious enhancing incidental lesions were routinely followed up by
postoperative MRI within 2 weeks to ensure that the lesion had not been missed
during surgical resection.
|
|
|
|
When enhancing incidental lesions were not biopsied, follow-up MRI was scheduled after 6 months, and the patient and her primary care physician were informed about this intent. If the lesion had not yet vanished but rather remained constant, further follow-up examinations were scheduled at 12, 24, and 36 months after the first investigation. Lesions that had increased in size were surgically explored. Patients (n = 45) who, in spite of our recommendation, could not be followed up on MRI for different reasons (e.g., discomfort during the first examination, refusal by the referring physician) were at least checked on conventional imaging for a period of 2-5 years (mean, 3.8 years) to rule out a growing malignant lesion that would eventually have been visible on mammography or sonography. Eight lesions were completely lost to follow-up.
Statistical Analysis
Ninety-five percent confidence intervals (CIs) for the positive predictive
values were computed with StatXact Version 4 (Cytel Software, Cambridge, MA)
using the Pearson's confidence intervals.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Category 2
The three cases with benign-appearing enhancing incidental lesions could be
followed up only on mammography and sonography. After 2 years 6 months, no
evidence of malignancy was obtained.
Category 3
Of the 112 enhancing incidental lesions classified as probably benign, 14
were histologically examined, and all were benign (five fibroadenomas, seven
foci of fibrocystic disease, one chronic inflammatory lesion, and one
cholesterol granuloma). Sixty-five lesions were followed up on MRI. Half of
these remained unchanged within a follow-up period of at least 24 months
(n = 32) and half disappeared (n = 33). One of the patients
was diagnosed with a contralateral carcinoma. Thirty patients were followed up
by conventional techniques, and no evidence of malignancy was seen during a
follow-up period of 2-5 years (mean, 3.8 years). Three patients in this
category were lost to follow-up.
Category 4a
Of 101 enhancing incidental lesions classified as indeterminate, 46 were
immediately studied by histology: 11 (24%) were malignant and 35 (76%) were
benign (four fibroadenomas, three papillomas, one hemangioma, two inflammatory
lesions, one radial scar, and 24 foci of fibrocystic disease). One lesion was
still present on postoperative MRI and remained unchanged for 2 years.
Thirty-nine lesions were followed up on MRI: 23 (60%) remained unchanged, 14
(35%) were no longer seen, and two (5%) had increased in size. These two
latter lesions were histologically examined and found to be invasive
carcinomas (diameters, 5 and 6 mm). Twelve patients could be followed up by
conventional techniques only, and no evidence of malignancy was obtained after
a follow-up period of at least 2 years (mean, 3.6 years). Three patients in
this category were lost to follow-up.
Category 4b
Of 45 enhancing incidental lesions classified as probably malignant, 41
were histologically examined: 31 lesions (76%) were malignant and 10 were
benign (one lobular carcinoma in situ, two fibroadenomas, one intramammary
lymph node, one papilloma, one inflammatory lesion, and four foci of
fibrocystic change [three of which exhibited atypia]). Three lesions were not
histologically examined but disappeared after neoadjuvant chemotherapy. One
patient refused any treatment. Her lesion remained unchanged during 30-months
follow-up.
Category 5
Of 13 enhancing incidental lesions classified as malignant, 10 were excised
and diagnosed as malignant. One patient refused both treatment and follow-up
examinations. One lesion disappeared after neoadjuvant chemotherapy. In one
mastectomy specimen, the enhancing incidental lesion (diameter, 5 mm) was not
specifically localized and thus not histologically examined.
Positive Predictive Value of Incidentally Detected Enhancing
Lesions
Only those lesions were entered into the calculation of the positive
predictive value that either were histologically diagnosed, disappeared
without chemotherapy, or remained constant on follow-up MRI examinations
within at least 2 years. The positive predictive value is 0 for category 3
(0/79; 95% CI, 0-5%), 15% for category 4a (13/86; 95% CI, 8-24%), 74% for
category 4b (31/42; 95% CI, 58-86%), and 100% for category 5 (10/10; 95% CI,
69-100%). The positive predictive value for the categories 4a, 4b, and 5
summarized into one malignant group is 39% (54/138; 95% CI, 31-48%).
Analysis of Malignant Lesions Identified Solely on MRI
Prevalence.—Of the 113 histologically examined enhancing
incidental lesions, 54 (48%) were malignant. Thus, at least 20% (54/274) of
all enhancing incidental lesions were malignant (not counting the four lesions
that disappeared under neoadjuvant chemotherapy). Of 508 malignant lesions
diagnosed in the 1273 patients, 346 had already been identified on
mammography. Through the combined use of mammography and sonography, in
addition, 79 lesions (16%, 79/508) were identified. The supplemental use of
MRI yielded another 54 lesions (11%, 54/508). Finally, 29 malignant lesions
(6%, 29/508) were detected on neither mammography, sonography, or MRI but were
only diagnosed as incidental lesions at histologic examination.
Histology.—Of the 54 malignant lesions detected solely on MRI, 40 were invasive ductal carcinomas, seven were tubular carcinomas, two were invasive lobular carcinomas, and five were DCIS. Twenty-seven were solitary lesions; 20, multicentric lesions; and seven, multifocal lesions.
Size.—The size of the malignant lesions ranged from 3-21 mm (mean, 7.1 mm; median, 6 mm) in 51 lesions. An exact size measurement was not obtained in a single diffusely growing invasive T3 carcinoma and in two DCIS lesions [11].
Reevaluation of mammograms and sonograms (Figs.
3A,
3B,
3C and
4).—With the exact
topography obtained on MRI in hand, we found 31 malignant lesions
retrospectively on mammograms (n = 3) and sonograms (n =
30). Enhancing incidental lesions that were not retrospectively seen were
predominantly small lesions (
5 mm). An 8-mm lesion that was located in
the resection margin of a preceding lumpectomy cavity was not seen on
conventional techniques. Not integrated into the graph
(Fig. 4) are the three lesions
for which exact size measurements could not be obtained. None of these three
lesions was retrospectively seen on conventional studies. Of the 59
histologically benign enhancing incidental lesions, only 13 could be
identified on repeated sonography (nine fibroadenomas, two papillomas, one
inflammatory cyst, one lymph node). All enhancing incidental lesions that were
no longer seen on the follow-up MRI examinations (n = 47), and 49 of
the 57 unchanged lesions could not be reidentified with certainty.
|
|
|
|
Clinical Implication
MRI revealed 274 enhancing incidental lesions in 210 patients. One hundred
thirteen of these enhancing incidental lesions (41%) in 101 patients (48%)
were histologically examined. The original treatment concept was modified in
46 (3.6%) of 1273 patients with 52 of the 54 malignant lesions solely detected
on MRI. In two patients, mastectomy had been originally scheduled. Fifty-nine
histologically proven benign enhancing incidental lesions in 54 patients (4.2%
of all patients) resulted in six core needle biopsies, 15 extensions of
incisions, nine additional ipsilateral incisions, nine additional
contralateral additional incisions, and seven surgical interventions that
would not have been undertaken without the diagnosis of enhancing incidental
lesions.
On the other hand, 153 of the 1273 patients were spared a biopsy because MRI helped to clarify indeterminate findings on clinical examination, sonography, and mammography. None of these lesions were found to be malignant after a follow-up period of at least 24 months. Besides the biopsies, we found that enhancing incidental lesions revealed on initial MRI resulted in consecutive follow-up examinations on MRI in 160 patients (12.6%).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The difference in prevalence of enhancing incidental lesions on MRI
reported in the literature may, in part, be explained by a nonstandardized
subjective definition of enhancing incidental lesions. In our study, we did
not include as enhancing incidental lesions disseminated dotlike, diffuse, or
patchy bilateral enhancing foci with a continuous signal-enhancement increase,
which are typically associated with hormonal stimulation. However, a fraction
of the enhancing incidental lesions in our study may have been hormone-induced
findings because they were no longer seen on the follow-up examination.
Another fact that should be considered is the advanced age of the
predominantly postmenopausal patients entered in our study (60% 
50
years). Finally, differences in technical aspects (field strength, temporal
and spatial resolution, dose of contrast media) and in previous diagnostic
modalities (solely mammography or in combination with sonography) must be
considered as well as the spectrum of indications for MRI.
In our study, the same evaluation criteria that applied to other enhancing lesions applied to enhancing incidental lesions. No enhancing incidental lesion that scored benign or probably benign turned out to be malignant. Within the indefinite group, 13% of the enhancing incidental lesions were malignant, and the positive predictive value increased to 74% and 100% in the probably malignant and malignant groups, respectively. These results are limited by the fact that not all lesions could be confirmed histologically but, in part, were followed up, only. However, ethical reasons prevented an integration of mandatory histologic examination of all enhancing incidental lesions.
On the other hand, we assume that the chance of malignancy is small if lesions remain constant during a period of 2-5 years or disappear. The positive predictive values calculated from our data compare well with the values found in the literature [3, 20-22]. Kuhl et al. [18, 19] also classified most enhancing incidental lesions in healthy patients as benign without any diagnostic dilemma.
In our study, 20% (54/274) of the enhancing incidental lesions were histologically confirmed to be malignant. This rate may be even slightly higher because six suspicious lesions were not histologically examined. Schmelzel et al. (RSNA meeting, November 2001) found a similar malignancy rate (21%, 29/137). In contrast, Brown et al. [9] found only one of 30 enhancing incidental lesions to be malignant. Reasons for this discrepancy might be the small fraction of histologically examined enhancing incidental lesions in their study (7/103 patients with one or more enhancing incidental lesions), the restriction of including patients with equivocal mammographic findings, and the short follow-up interval in some patients (2-42 months) usually not including MRI.
In the present study, two malignancies recognized by growth in the 6-month follow-up MRI would still have been missed on mammography at this time. Likewise, the malignancy rate of enhancing incidental lesions in the indication group of equivocal finding in conventional imaging was relatively low (7%, 7/105) in our study compared with other indication groups in which the values reached up to 40% (Table 1). Remarkable is the high malignancy rate of enhancing incidental lesions in the indication category of suspicious microcalcifications, suggesting that microcalcifications might represent a marker of increased risk. This indication was originally included to test the ability of MRI to rule out or confirm DCIS. All enhancing incidental lesions in this indication group were located distant from the suspicious microcalcifications or even in the contralateral breast. However, no such conclusion was drawn in a similar study by Westerhof et al. [23].
In our study, 11% of the histologically confirmed malignant lesions in the 1273 study patients (54/508) was found solely on MRI. Statistically, one or more as yet undiagnosed malignant foci were found in every 26th MRI examination (3.8%, 48/1273). Not only the malignancy rate of enhancing incidental lesions but also the rate of malignant lesions found on MRI depended on the indication for the study (Table 1): the highest rates were found when MRI was performed because of bloody secretions (8.7%), microcalcifications (8.4%), and searches for additional tumor foci in women with findings highly suspicious for malignancy (7.6%). This latter indication is one of the best explored aims of breast MRI in the literature [24-29]. Here, the examination rate with malignant lesions detected only on MRI reached up to 37% [26], depending on the prevalence of multifocal or multicentric tumor foci as well as the method of conventional diagnostics. As a consequence, this rate is higher in studies in which only mammography was used before MRI [26, 28] and is also reflected in a study by Berg and Gilbreath [30], in which the additional use of sonography detected six (14%) of 64 malignant foci in 40 patients that had not been seen on mammography alone. This rate corresponds well to the one of our study in which the supplementary use of sonography resulted in 79 additional malignant lesions (16%, 79/508) compared with 346 malignant lesions (68%, 346/508) that had been seen on mammography before. Based on these numbers, the benefit of MRI was an increase of detected malignancies by 25% over mammography alone.
The mean diameter of the 54 malignant lesions found solely on MRI (7.1 mm) was much smaller than the mean diameter (16 mm) of all malignant lesions (n = 508) found in the 1273 study patients. The mean sizes of malignancies detected solely on MRI as reported in the literature reaches from 7.6 to 16 mm [31-34].
Because MRI-guided localization and biopsy are difficult, expensive, and time-consuming, we have tried to reidentify enhancing incidental lesions on mammography and sonography, knowing their position, size, and shape from the MRIs. This identification was successful predominately on sonography in 31 (57%) of 54 of malignant lesions (Figs. 3A, 3B, 3C and 4). All except two malignant lesions greater than 5 mm in diameter could be relocalized on conventional studies. These two exceptions were a 20-mm recurrent tumor nodule located behind a silicone prosthesis, and an 8-mm residual tumor nodule hidden in extensive postoperative changes. Although these two false-negative findings were considered unavoidable, the remaining 23 lesions that were seen on retrospective evaluation had been missed on the initial sonogram and mammogram. One third of all lesions 5 mm in diameter and all smaller lesions could not be relocated even on thorough review of conventional diagnostics. Likewise, a diffusely growing invasive and two intraductal carcinomas could not be correlated successfully.
These data compare to those reported by Obdeijn et al. [34], who reidentified on sonography all 13 malignant lesions first detected on MRI in 31 patients. These lesions had a mean diameter of 16 mm (range, 6-50 mm). Brenner and Rothman [35], who examined the ability of MRI to detect a primary site of malignancy in the breast of patients with known lymph node metastases and normal findings on mammograms, also successfully reidentified all four malignant lesions on sonography. These studies and our own data do not support the view of Fischer et al. [3, 31], who could not relocalize 14 enhancing incidental lesions in 685 patients and hence preferred primary MRI-guided localization or biopsy over repeated conventional diagnostics. In contrast to the malignant lesions, benign lesions could rarely be reindentified on mammography or sonography. A putative explanation is that they often reflect fibrocystic changes with focal epithelial proliferations without a macroscopically visible mass or else inconstant focal accentuated changes caused by hormones.
When compared with mammography and sonography, MRI of the breast is complex and expensive. Both the referring physician and the patient expect an improved and more precise documentation or exclusion of malignancies— and no additional uncertainties. Unnecessary biopsies are a burden for the patient and the budget, and MRI follow-up examinations result in additional costs.
However, our study indicates that the benefit of MRI will likely outweigh the problems: on one side, MRI resulted in unnecessary surgical interventions (0.5%, 7/1273) and core biopsies (0.5%, 6/1273). Furthermore, independently scheduled surgical interventions were extended (3.2%, 41/1273), and MRI follow-up studies were performed (12.6%, 160/1273) without confirmation of suspicious findings. On the other side, malignant lesions that would otherwise have been missed could be identified in 3.6% (46/1273) of patients, and 12% (153/1273) of patients were spared invasive diagnostic procedures because findings in MRI studies were entirely negative. Eleven percent of the malignant lesions (54/508) were only diagnosed on MRI. In a report by Fischer et al. [3] on preoperative MRI of the breast in 463 patients (400 of whom had lesions that were suspicious for malignancy), the surgical approach was efficiently modified because of MRI findings in 66 patients (14.3%), whereas only 16 unnecessary additional interventions (3.5%) were undergone because of false-positive findings on MRI. Ten percent (39/405) of the malignant lesions, not counting multifocal lesions, were seen only on MRI,
On the basis of the findings of the present study, we propose the following management concept for enhancing incidental lesions and hence hope to further improve the efficacy of MRI in breast cancer detection (Fig. 5). All lesions are rated using the scoring system that was adapted to our technique. For enhancing incidental lesions that are considered probably malignant or malignant (categories 4b, 5) and are 5 mm or smaller, we recommend surgical resection after MRI-guided localization. Alternatively, MRI-guided vacuum biopsy is suggested where available. In our experience, MR-guided core needle biopsy is not sufficiently precise to safely target lesions that are 5 mm in diameter or less. For larger suspicious lesions, we try to reidentify them on sonography and mammography if diffuse growth or other circumstances do not render a successful correlation unlikely. If the lesion can be relocalized, sonographically or mammographically guided core biopsy or preoperative wire placement is performed. Only if no definite correlate is found, an MRI-guided biopsy is performed. Because the chance of finding malignant lesions in this subgroup is small, minimally invasive approaches should be preferred.
|
Patients with probably benign or benign-appearing enhancing incidental lesions are followed up. We did not observe any malignant lesion in this group. Therefore, it may be questioned if expensive MRI is economically justified or whether a close follow-up using conventional imaging for 2 years would be adequate to reliably exclude the small risk of a growing malignancy. In the subgroup of indefinite enhancing incidental lesions (category 4a), we decide on the further steps that are required on the basis of the findings of conventional diagnosis, provided the lesions can be reidentified. If not, MR-guided minimally invasive biopsy is considered the most elegant approach to rule out malignancy. In the subgroup of indefinite enhancing incidental lesions of less than 5 mm, we found only 4% (2/50) malignant lesions during follow-up, and even after an increase in size that was used as a criterion of malignancy, these lesions could still be removed at an early still mammographically occult and prognostically excellent stage.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Abe, R. A. Schmidt, R. N. Shah, A. Shimauchi, K. Kulkarni, C. A. Sennett, and G. M. Newstead MR-Directed ("Second-Look") Ultrasound Examination for Breast Lesions Detected Initially on MRI: MR and Sonographic Findings Am. J. Roentgenol., February 1, 2010; 194(2): 370 - 377. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Moy, K. Elias, V. Patel, J. Lee, J. S. Babb, H. K. Toth, and C. L. Mercado Is Breast MRI Helpful in the Evaluation of Inconclusive Mammographic Findings? Am. J. Roentgenol., October 1, 2009; 193(4): 986 - 993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nakano, M. Yoshida, K. Fujii, K. Yorozuya, Y. Mouri, J. Kousaka, T. Fukutomi, J. Kimura, T. Ishiguchi, K. Ohno, et al. Fusion of MRI and Sonography Image for Breast Cancer Evaluation Using Real-time Virtual Sonography with Magnetic Navigation: First Experience Jpn. J. Clin. Oncol., September 1, 2009; 39(9): 552 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. A. Berg, L. Gutierrez, M. S. NessAiver, W. B. Carter, M. Bhargavan, R. S. Lewis, and O. B. Ioffe Diagnostic Accuracy of Mammography, Clinical Examination, US, and MR Imaging in Preoperative Assessment of Breast Cancer Radiology, December 1, 2004; 233(3): 830 - 849. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
