10.1245/ASO.2004.12.008
Annals of Surgical Oncology 11:756-761 (2004)
© 2004 Society of Surgical Oncology
Preoperative Magnetic Resonance Imaging Evaluation for Breast Cancers After Sonographically Guided Core-Needle Biopsy: A Comparison Study
Yun-Chung Cheung, MD,
Yung-Liang Wan, MD,
Yung-Feng Lo, MD,
Wai-Man Leung, MD,
Shin-Cheh Chen, MD and
Swei Hsueh, MD
From the Department of Diagnostic Radiology (Y-CC, Y-LW), Chang Gung Memorial Hospital, College of Medicine, and School of Medical Technology; and Departments of Surgery (Y-FL, S-CC), Radiation Oncology (W-ML), and Pathology (SH), Chang Gung Memorial Hospital, Kwei Shan, Tao Yuan Hsien, Taiwan.
Correspondence: Address correspondence and reprint requests to: Shin-Cheh Chen, MD, Department of Surgery, Chang Gung Memorial Hospital, 5 Fu Hsing Road, Kwei Shan, Tao Yuan Hsien, Taiwan; Fax: 886-3-3330365; E-mail: alex2143{at}ms33.hinet.net
 |
ABSTRACT
|
|---|
Background: The aim of the study was to evaluate the efficacy of contrast-enhanced magnetic resonance imaging (MRI) for preoperative assessment of palpable breast cancer after sonographically guided percutaneous core-needle biopsy.
Methods: Thirty-six breast cancers in 35 women that had been diagnosed by sonographically guided core-needle biopsy prior to subsequent MRI were evaluated in this retrospective study. Radiological and pathological reports, multiplicity, retroareolar involvement, and the size of the breast cancers were reviewed. The cancer sizes, as derived from sonography and enhanced MRI, were correlated with histological size in greatest diameter by means of Pearson’s correlation. The threshold value for significance was set at P < .05.
Results: Synchronous breast cancers were revealed in the index cases by means of enhanced MRI (10), sonography (8), and mammography (7). Two of the 36 index cancers (5.6%) benefited from MRI assessment. Retroareolar cancer extension was observed with enhanced MRI in five index cancers. Of these, one was also noted on both a sonogram and a mammogram. Four of the index cancers (11.1%) benefited from the enhanced MRI. Overall, five index cancers (13.9%) benefited from the enhanced MRI. With a gold standard of histology, the mean cancer sizes were underestimated by sonography and overestimated by enhanced MRI. In comparison with sonography, a stronger association was noted between MRI and histological measurements, with coefficients of 0.657 and 0.882, respectively (P < .001).
Conclusions: In a clinical setting, MRI for preoperative assessment of breast cancers is warranted. Minimally invasive, percutaneous core-needle biopsy did not alter the clinical efficacy of the MRI evaluation.
Key Words: Breast cancer • Conservative mastectomy • Magnetic resonance imaging • Needle biopsy • Preoperative planning • Sonography
|
INTRODUCTION
|
|---|
Sensitivities of 93% to 100% have been documented for contrast-enhanced magnetic resonance imaging (MRI) in the detection of breast cancer. Unfortunately, the specificity range varies widely (29–98%), depending on technique and patient selection.1,2 Morphological features and the intensity-time curve of the breast mass demonstrated by enhanced MRI can help to differentiate between the benign and malignant variants; however, accuracy is not guaranteed.3,4 Pathological diagnosis is impossible as well. This is one of the reasons why MRI is not utilized for every patient presenting with a palpable breast mass. The expense and time required for MRI examinations are other significant limitations, but the benefits of preoperative planning outweigh these to some extent. The cost-efficacy of MRI in clinical practice has thus been the subject of continuing discussion. In view of these efficiencies, enhanced MRI for preoperative cancer planning should ideally be performed after a definitive pathological diagnosis has been delivered.
Sonography is a valuable adjunctive modality, essential to fulfillment of the morphological criteria required for differential diagnosis of a palpable breast mass. Sonograms enable faithful distinction of the nature of a cyst or solid tumor.5,6 For certain solid tumors, however, histological differentiation of the benign and malignant variants is difficult. Sonographically guided core-needle biopsy is usually recommended for histological diagnosis before selection of management. For breast-cancer patients wishing to undergo conservative mastectomy, the multiplicity, extension, and size of breast cancers are important factors in preoperative planning. Over the past few years, sonography and enhanced MRI have proved to be important adjunctive modalities for achievement of this preliminary assessment. However, comparison of the effectiveness of sonography and enhanced MRI in postbiopsy evaluation of breast cancer has rarely been discussed. In this study, therefore, the efficacy of preoperative MRI assessment of palpable breast cancers, previously diagnosed by percutaneous sonographically guided core-needle biopsy, was assessed and analyzed.
|
MATERIALS AND METHODS
|
|---|
Patients
Medical records for patients who underwent a total of 468 breast MRI examinations in our hospital from July 1998 to December 2002 were retrospectively reviewed. Forty of the patients with palpable breast masses had breast cancers diagnosed by sonographically guided core-needle biopsy prior to the breast MRI examination. Of these, 35 subjects subsequently underwent surgery at our hospital and 5 refused treatment at our hospital; mammographic examination was performed for all patients (26 at our institution and 9 at other hospitals). No fatty breasts were reported. Bilateral sonographic evaluations of whole breasts were performed to reveal features of the palpable masses and for preoperative assessment, including detection of synchronous masses and the extension and size of the index masses. After obtaining consent from the patients, we performed sonographically guided core-needle biopsy on the palpable index breast masses during the same examination session. In cases of breast cancer diagnosed by needle biopsy, MRI examination was subsequently arranged and performed within the following week, prior to the surgery.
Sonography and MRI Examination
A radiologist who specializes in breast imaging (CYC, who has greater than 5 years’ experience in sonography and sonographically guided core-needle biopsy and has done more than 300 breast MRI evaluations) performed the sonographic and MRI examinations in all cases. The multiplicity of the breast cancers, including multifocality (more than one focus in the same quadrant) and multicentricity (lesions involving more than one quadrant), was assessed. Regardless of whether the malignancies were multifocal or multicentric, all were defined as synchronous cancer on the basis of the criterion of being isolated cancer independent of the index breast cancer (i.e., the target of the sonographically guided core-needle biopsy), separated by at least 1 cm. Additionally, the extension of the breast cancer was evaluated, and the size of the breast cancer was measured in greatest diameter.
Sonography with a broadband linear-array transducer (5
9 MHz; Elegra; Siemens, Issaquah, WA) was used for bilateral whole-breast examination, with axial and sagittal arrays used in all cases. The index cancers were carefully evaluated from radial and antiradial orientations. After recording the sonographic features and sizes of the index breast cancers, as well as determining the presence of synchronous tumors, we performed sonographically guided core-needle biopsy, using a powerful automatic spring-loaded biopsy gun (Magnum; C. R. Bard, Covington, GA) and an 18-gauge cutting biopsy needle (C. R. Bard) with a 22-mm throw. The specimens were then stored in formalin solution and sent for histopathological examination.
After the histological diagnoses of breast cancer, all patients underwent dynamic enhanced MRI examination with a 1.5-T MR scanner with dedicated bilateral breast coils (MagnetomVision; Siemens, Erlangen, Germany) while in the prone position to ensure good imaging intensity. The image-acquisition protocol, including the pulse sequence and the number, thickness, and location of slices, was kept the same. Cancer detection was based on dynamic enhanced MRI, with use of the three-dimensional turbo fast low-angle shot (FLASH) pulse sequence in an axial plane. Both breasts were examined simultaneously. Dynamic imaging was performed with 14 continuously repeated sequences, including four precontrast and 10 postcontrast dynamic acquisitions. The dose of contrast agent (Magnevist; Berlex, Montville, NJ) was based on body weight (0.1 mmol/kg). The injection was commenced after completion of the four precontrast acquisitions at a constant rate, finishing at precisely 15 seconds for all patients.
Demonstration of index and synchronous breast cancers was based on the raw image data, as well as subtraction of the images at 2 and 4 minutes after gadolinium injection. Discrete enhanced masses, with patch or linear enhancement extending to 1 cm beneath the nipple-areolar region, were classified as retroareolar involvement. The cancer size was measured from the greatest diameter on the axial plane of the enhanced images at approximately 2 minutes. The multiplicity and retroareolar extension of the breast cancers were assessed.
Histological Correlation
Cancer size, presence of synchronous cancers, and cancer involvement were based on official reports from the pathology department. The histological cancer size was measured in the long axis of the mass. Regardless of whether the location was multifocal or multicentric, a synchronous breast cancer was defined as any other malignancy completely separate from the index growth.
Statistical Analysis
Our data were retrospectively analyzed on the basis of official reports from the radiology and pathology departments. The presence of synchronous cancers and the extent of retroareolar involvement were recorded from the mammograms, sonograms, and MRI. The mean cancer sizes, as measured with sonography, enhanced MRI, and histology, were calculated and analyzed with paired t-tests. Additionally, the sizes of the malignancies, as measured with sonography and MRI, were individually correlated with the measurements from pathology by means of Pearson’s correlation, with statistical significance set at P < .05.
|
RESULTS
|
|---|
In this clinical setting, 35 women (age range, 28–67 years; average, 43.8 years) with 36 palpable breast cancers were enrolled. One patient had bilateral breast carcinoma. Final histological diagnoses were invasive ductal carcinoma (n = 24), ductal carcinoma in situ (n = 5), medullary carcinoma (n = 2), tubular carcinoma (n = 2), mucinous carcinoma (n = 1), invasive lobular carcinoma (n = 1), and mixed ductal and lobular carcinoma in situ (n = 1). Sixteen cancers were detected in the right breast and 20 in the left. Modified radical mastectomy was used to excise 21 of the cancers, and conservative mastectomy was performed for 15. Axilla lymph node sampling or dissection was undertaken in all cases, with metastasis revealed in 13.
Multifocal/multicentric and Retroareolar Extension of Breast Cancer
Of these 36 index cancers, synchronous breast cancers were detected by mammography for 7 (19.4%), sonography for 8 (22.2%), and contrast-enhanced MRI for 10 (27.8%). The histological subtypes of the index breast cancers with synchronous malignancies, as observed on sonogram, were invasive ductal carcinoma (n = 4), ductal carcinoma in situ (n = 3), and mixed invasive ductal and lobular carcinoma in situ (n = 1). Two of the 10 index breast cancers with synchronous cancer (20%) were additionally detected on enhanced MRI. The histological subtypes of these two cancers were invasive ductal carcinoma and tubular carcinoma. However, the overall percentage of synchronous breast cancers detected only by enhanced MRI was just 5.6% of all the index cancers.
Retroareolar extension was noted in five index breast cancers by means of enhanced MRI. Of these, one was also demonstrated on both a sonogram and a mammogram. On the enhanced MRI, the cancer extended to the retroareolar region, revealing linear enhancement (n = 2; Fig. 1), patchy enhancement (n = 2; Fig. 2), and discrete masses (n = 1; Fig. 3). Of the 36 index cancers, the MRI examination was of significant benefit for four (11.1%). The histological subtypes of these four cancers were ductal carcinoma in situ (n = 2), invasive ductal carcinoma (n = 1), and invasive lobular carcinoma (n = 1). Of these, nipple involvement was proven histologically for two.
Received for publication December 3, 2003.
Accepted for publication April 21, 2004.
TOP
ABSTRACT
INTRODUCTION
). 
