- Email: [email protected]
Magnetic resonance imaging for preoperative evaluation of breast cancer: a comparative study with mammography and ultrasonography
Magnetic Resonance Imaging for Preoperative Evaluation of Breast Cancer: A Comparative Study with Mammography and Ultrasonography Tunetake Hata, MD, Hiromasa Takahashi, MD, PhD, Kenichi Watanabe, MD, PhD, Masato Takahashi, MD, PhD, Kazunori Taguchi, MD, PhD, Tomoo Itoh, MD, PhD, Satoru Todo, MD, PhD, FACS The widespread use of mammographic screening has led to increased detection of small tumors that are often difficult to diagnose with conventional imaging modalities such as mammography and ultrasonography. Intraductal spread of breast cancer, a principle risk factor for local recurrence, is also difficult to diagnose with mammography and ultrasonography. We investigated the clinical usefulness of magnetic resonance imaging of the breast in the therapy of breast cancer and we compared it with mammography and ultrasonography. STUDY DESIGN: A total of 183 patients with primary breast cancer underwent surgery at our institute between September 1, 1999, and November 30, 2002. They were examined preoperatively with magnetic resonance imaging, mammography, and ultrasonography. Magnetic resonance imaging evaluation included contrast-enhanced dynamic studies using IV injection of gadoliniumdiethylenetriamine pentaacetic acid. RESULTS: Detection rates of breast cancers by magnetic resonance imaging, mammography, and ultrasonography were 93.7%, 84.6%, and 97.3%, respectively (magnetic resonance imaging versus mammography, p ⬍ 0.05). Patterns of time-intensity curves in dynamic magnetic resonance imaging differed with histologic types. Sensitivity, specificity, and accuracy of detection of intraductal spread were 66.7%, 64.2%, and 65.6% with MRI; 22.2%, 85.7%, and 50% with mammography; and 20.6%, 85.2%, and 50% with ultrasonography, respectively (sensitivity, specificity, and accuracy; p ⬍ 0.05, respectively). CONCLUSIONS: Magnetic resonance imaging can diagnose breast cancer as accurately as ultrasonography and more accurately than mammography. Patterns of time-intensity curves correlated with tumor histology. In addition, magnetic resonance imaging can detect intraductal spread more accurately than the other two methods. Magnetic resonance imaging appears to be indispensable in breast-conserving surgery to minimize local recurrence. ( J Am Coll Surg 2004;198:190–197. © 2004 by the American College of Surgeons) BACKGROUND:
The widespread use of mammographic screening for breast cancer has led to increased detection of small tumors. With conventional breast imaging techniques, such as mammography (MMG) and ultrasonography (US), it is often difficult to determine whether these small tumors are malignant or benign. Magnetic reso-
nance imaging (MRI) was developed in the late 1980s1 and has been used for breast cancer diagnosis as a supplement to MMG and US. Its clinical usefulness for preoperative diagnosis of breast cancer has been controversial because of its costliness. Given that residual intraductal spread of breast cancer in the margin of conserved breast is a principal risk factor for local recurrence after breast-conserving surgery,2,3 preoperative detection of intraductal spread is extremely important. Some authors4-12 have reported that MRI can identify intraductal spread of breast cancer. To our knowledge, few consecutive studies of a large number of breast cancer cases have been performed comparing the efficacy of MRI with that of MMG and US.
No competing interests declared.
Received May 21, 2003; Revised October 8, 2003; Accepted October 13, 2003. From the First Department of Surgery (Hata, H Takahashi, Watanabe, M Takahashi, Taguchi, Todo) and Department of Pathology (Itoh), Hokkaido University, Sapporo, Japan. Correspondence address: Hiromasa Takahashi, MD, PhD, Kita-15 Nishi-7 Kita-ku, 060-8638 Sapporo, Japan.
© 2004 by the American College of Surgeons Published by Elsevier Inc.
190
ISSN 1072-7515/04/$30.00 doi:10.1016/j.jamcollsurg.2003.10.008
Vol. 198, No. 2, February 2004
In this study, we compared the clinical usefulness of MRI, MMG, and US in the diagnosis and management of breast cancer. Of special interest was the diagnostic potential of MRI with regard to breast cancer and intraductal spread. METHODS Between September 1, 1999, and November 30, 2002, 183 female patients with primary breast cancer underwent operations at the First Department of Surgery, Hokkaido University Hospital. Breast cancer was suspected in these patients on the basis of clinical examination, MMG, US, and fine-needle aspiration biopsy performed in outpatient clinics. We performed bilateral MRI preoperatively for close examination of the tumor and detection of intraductal spread of breast cancer that was defined as connected or disconnected lesions in mammary ducts around an invasive cancer. But in this series of patients, results of preoperative MRI study did not influence the decision of operation method. The indication of a breast-conserving operation was determined on the basis of clinical examination, MMG, and US. Patient demographics
Our 183 patients had 187 breast cancers, including 4 bilateral cancers. Patients ranged in age from 25 to 85 y (median, 53 y). A total of 129 patients visited our outpatient clinic being aware of palpable tumors and 54 patients’ tumors were detected in the screening of breast cancer without subjective symptoms, though 53 patients had palpable tumors. One patient was detected by MMG only. Clinical stages by TNM classification13 were: stage I, 71 tumors; stage II, 96 tumors; stage III, 15 tumors; and stage IV, 5 tumors. We performed modified radical mastectomy for 126 lesions and breastconserving operations for 61 lesions. Tumor size ranged from 0.5 to 12 cm (median, 1.9 cm). Sixteen tumors were ⬍ 1.0 cm, 97 tumors were between 1.0 and 2.0 cm, and 71 tumors were ⬎2.0 cm. Seven tumors were ductal carcinoma in situ (DCIS), 163 were invasive ductal carcinoma (IDC), 7 were invasive lobular carcinoma (ILC), 7 were mucinous carcinoma (MC), and 3 were others. MRI studies
MRI was performed on all 187 lesions with a 1.5 T Magnetom Symphony (Siemens), using a dedicated bi-
Hata et al
Usefulness of MRI Mammography
191
lateral breast coil with the patient in the prone position. First, T2-weighted axial images were acquired using the turbo spin echo method with 4,000/80 (repetition time milliseconds [ms]/echo time ms), a 251 ⫻ 256 matrix. The section thickness was 5 mm with 2.3 mm of slice gap. The field of view was 300 mm. Then, T1-weighted axial images were obtained using the spin echo method with 600/12 (repetition time ms/echo time ms), a 256 ⫻ 256 matrix. The section thickness was 5 mm with 2.3 mm of slice gap. The field of view was 300 mm. After rapid-bolus IV injection of gadoliniumdiethylenetriamine pentaacetic acid, 0.1 mmol/kg body weight, and a 20 mL saline solution flush, dynamic study was performed using 3D-VIBE in the affected breast. Six series were taken in the sagittal plane with 3.92/1.65 (repetition time ms/echo time ms), a flip angle of 12 degrees, and a 256 ⫻ 256 matrix. The section thickness was 2.55 mm without slice gap. The field of view was 300 mm. All the dynamic images were subtracted, and enhancement curves (time intensity curves) were drawn for regions of interest. The pattern of a timeintensity curve was defined as malignant when it showed a rapid enhancement pattern, with the intensity reaching a peak within 3 minutes, followed by a plateau or washout of the intensity.14 After the dynamic series, T1-weighted threedimensional images were obtained although suppressing the signal from fat, and maximum intensity projection images were reconstructed to produce a threedimensional view of the main tumor mass and areas of intraductal spread. Certified breast surgeons interpreted MRI images in consultation with certified radiologists. T2- and T1weighted images were used for the investigation of whole breast and positioning of the tumor. T1-weighed threedimensional images were not as clear as we could explore the main tumor and areas of intraductal spread because they were a composite of plane images. Consequently, criteria for the diagnosis of breast cancer were defined as follows: irregular shape and an internal heterogeneity of contrast enhancement or ringlike enhanced margins and a rapid enhancement pattern in the dynamic study, with a peak intensity reached within 3 minutes, as reported previously.15-17 Criteria for intraductal spread were defined as follows: a daughter tumor around the main tumor, a strandlike enhancement on the margin of the main tumor, or bridging enhancement between the
192
Hata et al
Usefulness of MRI Mammography
J Am Coll Surg
Figure 1. Magnetic resonance imaging of intraductal spread of breast cancer. Left, daughter tumors around main tumor. Right, a strandlike enhancement and bridging enhancement between the main tumor and the daughter tumor.
main tumor and the daughter tumor. MRI of intraductal spread of breast cancer is indicated in Figure 1.
largest tumor diameter defined the tumor size. Tumors were classified according to the histologic classification of the World Health Organization.18
Mammographic studies
Mammography (craniocaudal and mediolateral oblique views) was performed in 183 patients (with 182 lesions) using Senograph DMR (General Electronic) and was interpreted by certified breast surgeons. Criteria for malignancy were: high-density shadows with an irregular or spiculated shape, with or without minute pleomorphic or heterogeneous calcifications located as a group or in a segmental area. Intraductal spread of breast cancer was suggested by pleomorphic or heterogeneous and linear or branching calcifications in a linear and segmental area in the direction of the nipple. US studies
Ultrasonography was performed in 183 patients with 182 lesions using LOGIQ 400 with a 7.5-MHz linear array probe (General Electronic), and results were analyzed by certified breast surgeons. Criteria for malignancy were: an irregular shape, a rough border and low and heterogeneous internal echoes with irregular or thick boundary echoes, depth/width ratio ⬎1.0. Intraductal spread of breast cancer was indicated by a ductal dilatation connected with tumors. Histopathologic studies
Cross-sections of the resected specimens were made at 5-mm intervals perpendicular to the line connecting the nipple and the center of the tumor. Specimens were examined microscopically using hematoxylin and eosin staining. The size of the lesion was measured, and the
Statistical analyses
Rates of sensitivity, specificity, and accuracy were compared among MRI, MMG, and US using the chi-square test. The correlation of the extent of intraductal spread on MRI with the extent on pathologic examination was studied by calculating Pearson’s correlation coefficient. For all tests, a p value ⬍ 0.05 was considered to indicate statistical significance. Statistical analyses were performed using StatView version 5.0 (SAS Institute Inc). RESULTS Detection rates of breast cancers by MRI, MMG, and US were 93.7%, 84.6%, and 97.3%, respectively. The detection rate with MRI was significantly higher than with MMG (p ⬍ 0.05) and as high as US (Fig. 2). Tumors were classified by tumor size into three groups: ⬍1.0 cm, between 1.0 and 2.0 cm, and ⬎2.0 cm. Rates at which each technique detected tumors are shown in Figure 3. There was no significant difference between MRI and US. The detection rate with MMG was significantly lower than with MRI and US in tumors ⬍2.0 cm (p ⬍ 0.05). Rates at which the three methods detected tumors of various histologic types (ie, DCIS, IDC, ILC, and MC) are shown in Figure 4. In IDC, the detection rate with MMG was significantly lower than with MRI and US in tumors ⬍2.0 cm (p ⬍ 0.05). There was no significant difference among the three imaging methods in the other tumor types.
Vol. 198, No. 2, February 2004
Figure 2. Detection rates of breast cancers by means of magnetic resonance imaging (MRI), mammography (MMG), and ultrasonography (US). The detection rate with MRI was significantly higher than with MMG and as high as with US. *p ⬍ 0.05.
When the dynamic studies using gadoliniumdiethylenetriamine pentaacetic acid were performed,
Hata et al
Usefulness of MRI Mammography
193
malignant patterns were observed in 89.2% of IDC cases, 50.0% of MC cases, 42.9% of ILC cases, and 60.0% of DCIS cases. The rate of malignant patterns in IDC was significantly higher than for other histologic types (p ⫽ 0.0003) (Figure 5). Rates of sensitivity, specificity, and accuracy of detection of intraductal spread for each of the three techniques is shown in Figure 6. The sensitivity and accuracy of MRI were significantly superior to those of MMG and US (p ⬍ 0.05), though the specificity of MRI was significantly inferior to MMG and US (p ⬍ 0.05). In 54 of 68 cases in which intraductal spread was detected by MRI and the extent was determined pathologically, the correlation between the distance on MRI and pathologically measured distance about intraductal spread of breast cancer was analyzed. A significant correlation between the distance on MRI and pathologic distance was identified (r ⫽ 0.42, p ⫽ 0.0015) (Figure 7). The distance of intraductal spread measured on MRI was greater than the actual pathologic extent in 21 (38.9%), smaller in 26 (48.1%), and equal to the actual pathologic extent in 7 (13.0%) of cases. DISCUSSION In 1986, Heywang and colleagues1 reported that breast carcinoma showed more prominent and earlier enhancement than normal breast tissue when studied with MRI
Figure 3. Tumor size and detection rates of breast cancers with magnetic resonance imaging, black bar; mammography, spotted bar; and ultrasonography, striped bar. Tumors were classified into three groups of tumor size: ⬍1.0, ⱖ1.0, and ⬎2.0 cm. The detection rate with mammography was significantly lower than with magnetic resonance imaging and ultrasonography in tumors ⬍ 2.0 cm.
194
Hata et al
Usefulness of MRI Mammography
J Am Coll Surg
Figure 4. Histologic classification and detection rates of breast cancers with magnetic resonance imaging, black bar; mammography, spotted bar; and ultrasonography, striped bar. Tumors were classified into four histologic tumor types. The detection rate with mammography was significantly lower than with magnetic resonance imaging and ultrasonography in invasive ductal carcinoma.
using gadolinium-diethylenetriamine pentaacetic acid. Since then, contrast-enhanced MRI (ie, dynamic MRI) has developed as a promising modality and a useful supplement to MMG and US for detection of breast cancer, especially for detection of nonpalpable tumors19 and for detection of undetected tumors by MMG and US.20
Figure 5. Histologic classification and patterns of time-intensity curve. Tumors were classified into four histologic tumor types: mucinous carcinoma (MC); invasive lobular carcinoma (ILC); invasive ductal carcinoma (IDC); and ductal carcinoma in situ (DCIS). Rate of a malignant pattern in IDS was significantly higher than other histologic types, p ⫽ 0.0003. Black bar, malignant pattern; spotted bar, benign pattern.
Orel and colleagues,21 pointed out that there were no standard interpretation criteria for evaluating MRI studies, no consensus on what constituted clinically important enhancement, no clearly defined indications for the use of MRI, and uncertain cost-effectiveness. Dynamic MRI of the breast has a high degree of sensitivity for the diagnosis of breast cancer, with most studies reporting sensitivity in excess of 90%.19,22-24 Similarly, in our study, MRI could identify breast cancer in accordance with morphologic criteria confirmed to be useful by Wedegartner and colleagues.25 Its rate of detection was superior to that of MMG and identical to that of US. Malur and colleagues also reported that MRI had higher sensitivity for invasive ductal cancer than MMG and US, though MRI was not superior to a combination of MMG and US.10 In the literature, MRI is more accurate than MMG and US in determining invasive tumor size.4,20,26 In our study, detection rates of breast cancer with MRI and US were unchanged regardless of tumor size, whereas the detection rate with MMG fell as tumor size decreased, because diagnosis had to be made on the basis of morphologic criteria alone. With MRI, small tumors were diagnosed as well as large tumors, because diagnosis was based on a combination of morphologic criteria and the findings of the dynamic study. For a similar reason,
Vol. 198, No. 2, February 2004
Hata et al
Usefulness of MRI Mammography
195
Figure 6. Rate of sensitivity, specificity, and accuracy of detection of intraductal spread of breast cancer by means of magnetic resonance imaging (MRI), mammography (MMG), and ultrasonography (US). The sensitivity and accuracy of MRI was significantly superior to those of MMG and US, although the specificity of MRI was significantly inferior to MMG and US. Black bar, sensitivity; spotted bar, specificity; striped bar, accuracy. MRI versus MMG and US: sensitivity, p ⬍ 0.05; specificity, p ⬍ 0.05; accuracy, p ⬍ 0.05.
MRI could detect any histologic type of breast cancer, such as DCIS, IDC, ILC, and MC, although some cases of ductal carcinoma in situ or mucinous carcinoma were not detected on MMG or US. Ando and colleagues27 demonstrated that the pattern of tumor vascularity as shown by MRI also could help differentiate malignant tumors from benign lesions that were enhanced on contrast administration. In our study, breast surgeons who diagnosed breast cancer interpreted MRI images. Still, this is hardly considered to bias the results because we
Figure 7. Correlation between the extent of intraductal spread of breast cancer as measured by magnetic resonance imaging (MRI) and the pathologic measured distance. A significant correlation between the distance on MRI and pathologic distance was identified.
adopted highly objective criteria for the diagnosis of breast cancer. With respect to time-intensity curves in dynamic MRI of the breast, Kuhl and colleagues14 classified curve patterns as type I, which was a straight or curved time course with steady intensity increase; type II, a plateau of intensity with a sharp bend after the initial upstroke; or type III, a washout of intensity after the initial upstroke. According to their report, types I, II, and III were found in 83.0%, 11.5%, and 5.5% of benign solid tumors and fibrocystic changes, respectively, whereas types I, II, and III were found in 8.9%, 33.6%, and 57.4% of malignant lesions, respectively. Accordingly, in dynamic MRI of the breast, the pattern of the time-intensity curve is an important criterion in differentiating benign and malignant enhancing lesions. In our study, patterns of time intensity curve in dynamic MRI differed with histologic types. This finding is consistent with the fact that some malignant lesions, such as certain invasive ductal and lobular carcinomas and certain ductal carcinoma in situ lesions, are not enhanced rapidly, but their morphology nevertheless suggests the presence of malignancy.28,29 Kuhl and colleagues14 also found that lobular or scirrhous ductal invasive cancers could exhibit a benign pattern of time-intensity curve because of slow enhancement and poor angiogenic activity. About the diagnosis of breast cancer, we can summarize that MRI provides not only the morphologic diag-
196
Hata et al
Usefulness of MRI Mammography
nosis, but also the opportunity to evaluate vascularity with a contrast medium, which is impossible for other imaging modalities, such as MMG and US. With regard to intraductal spread of breast cancer, microscopic intraductal spread is often unrecognized during surgery and can cause local recurrence.2,3 Reported rates of local recurrence range from 5.5% to 27.9%.30-33 If local recurrence occurs after breastconserving surgery, the conserved breast is usually removed. Consequently, accurate assessment of the extent of intraductal spread is essential for determining the feasibility of breast-conserving surgery and the range of excision of the mammary gland. US and MMG have been used for evaluating the extent of breast cancer, but both have limitations. Realtime US must be performed by a highly skilled individual, and its interpretation is more complicated than that of MRI and MMG because fixed figures cannot be always shown. Satake and colleagues34 have reported that the sensitivity, specificity, and accuracy rates in the detection of intraductal spread with US were 89%, 76%, and 85%, respectively. In our study, the sensitivity rate of US for detecting intraductal spread was very low in comparison to MRI, probably because our criterion of intraductal spread on US was too simplified compared with that of Satake and colleagues.34 Given that MMG cannot identify intraductal spread that does not include minute calcifications, it was predictable that the sensitivity rate of MMG for detecting intraductal spread would be lower than MRI in our study. Satake and colleagues34 also reported that the sensitivity of MMG was lower than that of US and MRI. According to the recent literature, intraductal spread can be enhanced by contrast medium, thus making it detectable by dynamic MRI;35 the rate of the sensitivity ranges from 55% to 100%.4-9,10,35-38 Some authors have shown that MRI detects intraductal spread more accurately than MMG and US,10,20 as we indicated in our study. Recently, three-dimensional MRI has been used for detection of intraductal spread, and its usefulness beyond computed tomography or US has been reported.11,12 In 54 of 68 cases in which intraductal spread was detected by MRI and the extent was determined pathologically, a significant positive correlation between the distance on MRI and pathologic distance was identified. The distance of intraductal spread measured on MRI was greater than the actual pathologic extent in 38.9%
J Am Coll Surg
of cases. Fibrocystic disease around the intraductal spread and lymphocyte infiltration and fibrous proliferation associated with a tumor can lead to enhancement and overestimation of the distance on MRI, as other authors also have mentioned.6,19,36 Given that MRI overestimates the extent of intraductal spread, there is the potential risk of unnecessary mastectomies. This problem should be taken into account on the occasion of planning of breast-conserving operation with MRI. In conclusion, MRI can diagnose breast cancer as accurately as US and more accurately than MMG without respect to tumor size or histologic type. In addition, patterns of time-intensity curves may enable us to predict histologic type. MRI can detect intraductal spread of breast cancer more accurately than MMG and US. It is predictable that preoperative MRI reduces residual intraductal spread of breast cancer in the margin of conserved breast and the rate of local recurrence after breastconserving surgery. So, in view of the particular advantages of MRI in the diagnosis and preoperative characterization of breast cancer, MRI is considered indispensable for determining the appropriateness of breast-conserving surgery and for planning the surgical approach to prevent postoperative local recurrence. The clinical outcomes after operation planning with MRI of the breast remain unsolved. A further study is required to answer this question. Acknowledgment: We are deeply indebted to Akiko Tukahara, MD, at the Department of Radiology Hokkaido University Hospital, who consulted about the interpretation of magnetic resonance images.
REFERENCES 1. Heywang SH, Hahn D, Schmidt H, et al. MR imaging of the breast using gadolinium-DTPA. J Comput Assist Tomograr 1986;10:99–204. 2. Lindley R, Bulman A, Parsons P, et al. Histologic features predictive of an increased risk of early local recurrence after treatment of breast cancer by local tumor excision and radical radiotherapy. Surgery 1989;105:13–20. 3. Holland R, Connolly JL, Gelman R, et al. The presence of an extensive intraductal component following a limited excision correlates with prominent residual disease in the remainder of the breast. J Clin Oncol 1990;8:113–118. 4. Mumtaz H, Hall-Craggs MA, Davidson T, et al. Staging of symptomatic primary breast cancer with MR imaging. Am J Roentgenol 1997;169:417–424. 5. Hiramatu H, Enomoto K, Ikeda T, et al. The role of contrast-
Vol. 198, No. 2, February 2004
6.
7.
8.
9.
10.
11. 12. 13. 14.
15. 16. 17. 18. 19. 20. 21.
enhanced high resolution MRI in the surgical planning of breast cancer. Breast Cancer 1997;4:285–290. Date S. Diagnosis of intraductal spread of breast cancer by highresolution MR imaging: correlation between MR imaging and pathohistological findings. Nippon Igaku Hoshasen Gakkai Zasshi 1998;58:212–220. Drew PJ, Chatterjee S, Turnbull LW, et al. Dynamic contrast enhanced magnetic resonance imaging of the breast is superior to triple assessment for the pre-operative detection of multifocal breast cancer. Ann Surg Oncol 1999;6:599–603. Gatzemeier W, Liesch T, Stylianou A, et al. Preoperative MR mammography in breast carcinoma. Effect on operative treatment from the surgical viewpoint. Chirurg 1999;70:1460– 1468. Shiraishi A, Suzuki M, Nozu S, et al. Diagnosis of breast cancer extent and enhance patterns using 3D-dynamic imaging: correlation with intraductal component. Nippon Igaku Hoshasen Gakkai Zasshi 1999;59:122–130. Malur S, Wurdinger S, Moritz A, et al. Comparison of written reports of mammography, sonography and magnetic resonance mammography for preoperative evaluation of breast lesions, with special emphasis on magnetic resonance mammography. Breast Cancer Res 2001;3:55–60. Tamaki Y, Akashi-Tanaka S, Ishida T, et al. 3D imaging of intraductal spread of breast cancer and its clinical application for navigation surgery. Breast Cancer Res 2002;9:289–95. Nakahara H, Namba K, Wakamatsu H, et al. Extension of breast cancer: comparison of CT and MRI. Radiat Med 2002;20:17– 23. Hermanke LH, Sobin LH. International Union Against Cancer (UICC): TNM classification of malignant tumors, 4th ed. Berlin: Springer; 1989. Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR imaging. Are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 1989;211:101– 110. Heywang SH, Wolf A, Pruss E, et al. MR imaging of the breast with Gd-DTPA; use and limitations. Radiology 1989;171:95– 103. Kaiser WA, Zeitler E. MR imaging of the breast; Fast imaging sequences with and without Gd-DTPA. Preliminary observations. Radiology 1989;170:681–686. Stack JP, Redmond OM, Codd MB, et al. Breast disease: tissue characterization with Gd-DTPA enhancement profiles. Radiology 1990;174:491–494. WHO. Histological typing of breast tumours. International histological classification of tumours. 2nd ed. Geneva: World Health Organization; 1981. Gilles R, Guinebretiere JM, Lucidarme O, et al. Nonpalpable breast tumors: diagnosis with contrast-enhanced subtraction dynamic MR imaging. Radiology 1994;191:625–631. Boetes C, Mus RD, Holland R, et al. Breast tumor: comparative accuracy of MR imaging relative to mammography and US for demonstrating extent. Radiology 1995;197:743–747. Orel SG, Schnall MD. MR imaging of the breast for the detection, diagnosis, and staging of breast cancer. Radiology 2001; 220:13–30.
Hata et al
Usefulness of MRI Mammography
197
22. Bone B, Aspelin P, Bronge L, et al. Sensitivity and specificity of MR mammography with histopathological correlation in 250 breasts. Acta Radiol 1996;37:208–213. 23. Helbich TH, Becherer A, Trattnig S, et al. Differentiation of benign and malignant breast lesions: MR imaging versus Tc99m sestamibi scintimammography. Radiology 1997;202:421– 429. 24. Fischer U, Kopka L, Grabbe E. Breast carcinoma: effect of preoperative contrast-enhanced MR imaging on the therapeutic approach. Radiology 1999;213:881–888. 25. Wedegartner U, Bick U, Wortler K, et al. Differentiation between benign and malignant findings on MR-mammography: usefulness of morphological criteria. Eur Radiol 2001;11:1645– 1650. 26. Davis PL, Staiger MJ, Harris KB, et al. Breast cancer measurement with magnetic resonance imaging, ultrasonography, and mammography. Breast Cancer Res Treat 1996;37:1–9. 27. Ando Y, Fukatsu H, Ishiguchi T, et al. Diagnostic utility of tumor vascularity on magnetic resonance imaging of the breast. Magn Reson Imaging 2000;18:807–813. 28. Nunes LW, Schnall MD, Orel SG, et al. Breast MR imaging. Interpretation model. Radiology 1997;202:833–841. 29. Orel SG. Differentiating benign from malignant enhancing lesions identified at MR imaging of the breast: are time-signal intensity curves an accurate predictor? Radiology 1999;211: 5–7. 30. Abner AL, Recht A, Eberlein T, et al. Prognosis following salvage mastectomy for recurrence in the breast after conservative surgery and radiation therapy for early-stage breast cancer. J Clin Oncol 1993;11:44–48. 31. Fisher B, Bauer M, Margolese R, et al. Five-year results of a randomized clinical trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N Engl J Med 1985;14:312 665–673. 32. Lagios MD, Margolin FR, Westdahl PR, et al. Mammographically detected duct carcinoma in situ. Frequency of local recurrence following tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer 1989;63:15 618–624. 33. Fisher ER, Sass R, Fisher B, et al. Pathologic findings of the National Surgical Adjuvant Breast Project (Protocol 6). I. Intraductal carcinoma (DCIS). Cancer 1986;15(57):197–208. 34. Satake H, Shimamoto K, Sawaki A, et al. Role of ultrasonography in the detection of intraductal spread of breast cancer: correlation with pathologic findings, mammography and MR imaging. Eur Radiol 2000;10:1726–1732. 35. Gilles R, Zafrani B, Guinebretiere JM, et al. Ductal carcinoma in situ: MR imaging-histopathologic correlation. Radiology 1995;196:415–419. 36. Kawashima H. Fat suppressed dynamic MR imaging of breast cancer. Radiologic-pathologic correlation. Nippon Igaku Hoshasen Gakkai Zasshi 1996;56:459–569. 37. Orel SG, Mendonca MH, Reynolds C, et al. MR imaging of ductal carcinoma in situ. Radiology 1997;202:413–420. 38. Ando Y, Fukatsu H, Ishigaki T, et al. Intramammary extension of breast cancer: diagnosis with MR mammography. Breast Cancer 1998;5:291–300.
The widespread use of mammographic screening for breast cancer has led to increased detection of small tumors. With conventional breast imaging techniques, such as mammography (MMG) and ultrasonography (US), it is often difficult to determine whether these small tumors are malignant or benign. Magnetic reso-
nance imaging (MRI) was developed in the late 1980s1 and has been used for breast cancer diagnosis as a supplement to MMG and US. Its clinical usefulness for preoperative diagnosis of breast cancer has been controversial because of its costliness. Given that residual intraductal spread of breast cancer in the margin of conserved breast is a principal risk factor for local recurrence after breast-conserving surgery,2,3 preoperative detection of intraductal spread is extremely important. Some authors4-12 have reported that MRI can identify intraductal spread of breast cancer. To our knowledge, few consecutive studies of a large number of breast cancer cases have been performed comparing the efficacy of MRI with that of MMG and US.
No competing interests declared.
Received May 21, 2003; Revised October 8, 2003; Accepted October 13, 2003. From the First Department of Surgery (Hata, H Takahashi, Watanabe, M Takahashi, Taguchi, Todo) and Department of Pathology (Itoh), Hokkaido University, Sapporo, Japan. Correspondence address: Hiromasa Takahashi, MD, PhD, Kita-15 Nishi-7 Kita-ku, 060-8638 Sapporo, Japan.
© 2004 by the American College of Surgeons Published by Elsevier Inc.
190
ISSN 1072-7515/04/$30.00 doi:10.1016/j.jamcollsurg.2003.10.008
Vol. 198, No. 2, February 2004
In this study, we compared the clinical usefulness of MRI, MMG, and US in the diagnosis and management of breast cancer. Of special interest was the diagnostic potential of MRI with regard to breast cancer and intraductal spread. METHODS Between September 1, 1999, and November 30, 2002, 183 female patients with primary breast cancer underwent operations at the First Department of Surgery, Hokkaido University Hospital. Breast cancer was suspected in these patients on the basis of clinical examination, MMG, US, and fine-needle aspiration biopsy performed in outpatient clinics. We performed bilateral MRI preoperatively for close examination of the tumor and detection of intraductal spread of breast cancer that was defined as connected or disconnected lesions in mammary ducts around an invasive cancer. But in this series of patients, results of preoperative MRI study did not influence the decision of operation method. The indication of a breast-conserving operation was determined on the basis of clinical examination, MMG, and US. Patient demographics
Our 183 patients had 187 breast cancers, including 4 bilateral cancers. Patients ranged in age from 25 to 85 y (median, 53 y). A total of 129 patients visited our outpatient clinic being aware of palpable tumors and 54 patients’ tumors were detected in the screening of breast cancer without subjective symptoms, though 53 patients had palpable tumors. One patient was detected by MMG only. Clinical stages by TNM classification13 were: stage I, 71 tumors; stage II, 96 tumors; stage III, 15 tumors; and stage IV, 5 tumors. We performed modified radical mastectomy for 126 lesions and breastconserving operations for 61 lesions. Tumor size ranged from 0.5 to 12 cm (median, 1.9 cm). Sixteen tumors were ⬍ 1.0 cm, 97 tumors were between 1.0 and 2.0 cm, and 71 tumors were ⬎2.0 cm. Seven tumors were ductal carcinoma in situ (DCIS), 163 were invasive ductal carcinoma (IDC), 7 were invasive lobular carcinoma (ILC), 7 were mucinous carcinoma (MC), and 3 were others. MRI studies
MRI was performed on all 187 lesions with a 1.5 T Magnetom Symphony (Siemens), using a dedicated bi-
Hata et al
Usefulness of MRI Mammography
191
lateral breast coil with the patient in the prone position. First, T2-weighted axial images were acquired using the turbo spin echo method with 4,000/80 (repetition time milliseconds [ms]/echo time ms), a 251 ⫻ 256 matrix. The section thickness was 5 mm with 2.3 mm of slice gap. The field of view was 300 mm. Then, T1-weighted axial images were obtained using the spin echo method with 600/12 (repetition time ms/echo time ms), a 256 ⫻ 256 matrix. The section thickness was 5 mm with 2.3 mm of slice gap. The field of view was 300 mm. After rapid-bolus IV injection of gadoliniumdiethylenetriamine pentaacetic acid, 0.1 mmol/kg body weight, and a 20 mL saline solution flush, dynamic study was performed using 3D-VIBE in the affected breast. Six series were taken in the sagittal plane with 3.92/1.65 (repetition time ms/echo time ms), a flip angle of 12 degrees, and a 256 ⫻ 256 matrix. The section thickness was 2.55 mm without slice gap. The field of view was 300 mm. All the dynamic images were subtracted, and enhancement curves (time intensity curves) were drawn for regions of interest. The pattern of a timeintensity curve was defined as malignant when it showed a rapid enhancement pattern, with the intensity reaching a peak within 3 minutes, followed by a plateau or washout of the intensity.14 After the dynamic series, T1-weighted threedimensional images were obtained although suppressing the signal from fat, and maximum intensity projection images were reconstructed to produce a threedimensional view of the main tumor mass and areas of intraductal spread. Certified breast surgeons interpreted MRI images in consultation with certified radiologists. T2- and T1weighted images were used for the investigation of whole breast and positioning of the tumor. T1-weighed threedimensional images were not as clear as we could explore the main tumor and areas of intraductal spread because they were a composite of plane images. Consequently, criteria for the diagnosis of breast cancer were defined as follows: irregular shape and an internal heterogeneity of contrast enhancement or ringlike enhanced margins and a rapid enhancement pattern in the dynamic study, with a peak intensity reached within 3 minutes, as reported previously.15-17 Criteria for intraductal spread were defined as follows: a daughter tumor around the main tumor, a strandlike enhancement on the margin of the main tumor, or bridging enhancement between the
192
Hata et al
Usefulness of MRI Mammography
J Am Coll Surg
Figure 1. Magnetic resonance imaging of intraductal spread of breast cancer. Left, daughter tumors around main tumor. Right, a strandlike enhancement and bridging enhancement between the main tumor and the daughter tumor.
main tumor and the daughter tumor. MRI of intraductal spread of breast cancer is indicated in Figure 1.
largest tumor diameter defined the tumor size. Tumors were classified according to the histologic classification of the World Health Organization.18
Mammographic studies
Mammography (craniocaudal and mediolateral oblique views) was performed in 183 patients (with 182 lesions) using Senograph DMR (General Electronic) and was interpreted by certified breast surgeons. Criteria for malignancy were: high-density shadows with an irregular or spiculated shape, with or without minute pleomorphic or heterogeneous calcifications located as a group or in a segmental area. Intraductal spread of breast cancer was suggested by pleomorphic or heterogeneous and linear or branching calcifications in a linear and segmental area in the direction of the nipple. US studies
Ultrasonography was performed in 183 patients with 182 lesions using LOGIQ 400 with a 7.5-MHz linear array probe (General Electronic), and results were analyzed by certified breast surgeons. Criteria for malignancy were: an irregular shape, a rough border and low and heterogeneous internal echoes with irregular or thick boundary echoes, depth/width ratio ⬎1.0. Intraductal spread of breast cancer was indicated by a ductal dilatation connected with tumors. Histopathologic studies
Cross-sections of the resected specimens were made at 5-mm intervals perpendicular to the line connecting the nipple and the center of the tumor. Specimens were examined microscopically using hematoxylin and eosin staining. The size of the lesion was measured, and the
Statistical analyses
Rates of sensitivity, specificity, and accuracy were compared among MRI, MMG, and US using the chi-square test. The correlation of the extent of intraductal spread on MRI with the extent on pathologic examination was studied by calculating Pearson’s correlation coefficient. For all tests, a p value ⬍ 0.05 was considered to indicate statistical significance. Statistical analyses were performed using StatView version 5.0 (SAS Institute Inc). RESULTS Detection rates of breast cancers by MRI, MMG, and US were 93.7%, 84.6%, and 97.3%, respectively. The detection rate with MRI was significantly higher than with MMG (p ⬍ 0.05) and as high as US (Fig. 2). Tumors were classified by tumor size into three groups: ⬍1.0 cm, between 1.0 and 2.0 cm, and ⬎2.0 cm. Rates at which each technique detected tumors are shown in Figure 3. There was no significant difference between MRI and US. The detection rate with MMG was significantly lower than with MRI and US in tumors ⬍2.0 cm (p ⬍ 0.05). Rates at which the three methods detected tumors of various histologic types (ie, DCIS, IDC, ILC, and MC) are shown in Figure 4. In IDC, the detection rate with MMG was significantly lower than with MRI and US in tumors ⬍2.0 cm (p ⬍ 0.05). There was no significant difference among the three imaging methods in the other tumor types.
Vol. 198, No. 2, February 2004
Figure 2. Detection rates of breast cancers by means of magnetic resonance imaging (MRI), mammography (MMG), and ultrasonography (US). The detection rate with MRI was significantly higher than with MMG and as high as with US. *p ⬍ 0.05.
When the dynamic studies using gadoliniumdiethylenetriamine pentaacetic acid were performed,
Hata et al
Usefulness of MRI Mammography
193
malignant patterns were observed in 89.2% of IDC cases, 50.0% of MC cases, 42.9% of ILC cases, and 60.0% of DCIS cases. The rate of malignant patterns in IDC was significantly higher than for other histologic types (p ⫽ 0.0003) (Figure 5). Rates of sensitivity, specificity, and accuracy of detection of intraductal spread for each of the three techniques is shown in Figure 6. The sensitivity and accuracy of MRI were significantly superior to those of MMG and US (p ⬍ 0.05), though the specificity of MRI was significantly inferior to MMG and US (p ⬍ 0.05). In 54 of 68 cases in which intraductal spread was detected by MRI and the extent was determined pathologically, the correlation between the distance on MRI and pathologically measured distance about intraductal spread of breast cancer was analyzed. A significant correlation between the distance on MRI and pathologic distance was identified (r ⫽ 0.42, p ⫽ 0.0015) (Figure 7). The distance of intraductal spread measured on MRI was greater than the actual pathologic extent in 21 (38.9%), smaller in 26 (48.1%), and equal to the actual pathologic extent in 7 (13.0%) of cases. DISCUSSION In 1986, Heywang and colleagues1 reported that breast carcinoma showed more prominent and earlier enhancement than normal breast tissue when studied with MRI
Figure 3. Tumor size and detection rates of breast cancers with magnetic resonance imaging, black bar; mammography, spotted bar; and ultrasonography, striped bar. Tumors were classified into three groups of tumor size: ⬍1.0, ⱖ1.0, and ⬎2.0 cm. The detection rate with mammography was significantly lower than with magnetic resonance imaging and ultrasonography in tumors ⬍ 2.0 cm.
194
Hata et al
Usefulness of MRI Mammography
J Am Coll Surg
Figure 4. Histologic classification and detection rates of breast cancers with magnetic resonance imaging, black bar; mammography, spotted bar; and ultrasonography, striped bar. Tumors were classified into four histologic tumor types. The detection rate with mammography was significantly lower than with magnetic resonance imaging and ultrasonography in invasive ductal carcinoma.
using gadolinium-diethylenetriamine pentaacetic acid. Since then, contrast-enhanced MRI (ie, dynamic MRI) has developed as a promising modality and a useful supplement to MMG and US for detection of breast cancer, especially for detection of nonpalpable tumors19 and for detection of undetected tumors by MMG and US.20
Figure 5. Histologic classification and patterns of time-intensity curve. Tumors were classified into four histologic tumor types: mucinous carcinoma (MC); invasive lobular carcinoma (ILC); invasive ductal carcinoma (IDC); and ductal carcinoma in situ (DCIS). Rate of a malignant pattern in IDS was significantly higher than other histologic types, p ⫽ 0.0003. Black bar, malignant pattern; spotted bar, benign pattern.
Orel and colleagues,21 pointed out that there were no standard interpretation criteria for evaluating MRI studies, no consensus on what constituted clinically important enhancement, no clearly defined indications for the use of MRI, and uncertain cost-effectiveness. Dynamic MRI of the breast has a high degree of sensitivity for the diagnosis of breast cancer, with most studies reporting sensitivity in excess of 90%.19,22-24 Similarly, in our study, MRI could identify breast cancer in accordance with morphologic criteria confirmed to be useful by Wedegartner and colleagues.25 Its rate of detection was superior to that of MMG and identical to that of US. Malur and colleagues also reported that MRI had higher sensitivity for invasive ductal cancer than MMG and US, though MRI was not superior to a combination of MMG and US.10 In the literature, MRI is more accurate than MMG and US in determining invasive tumor size.4,20,26 In our study, detection rates of breast cancer with MRI and US were unchanged regardless of tumor size, whereas the detection rate with MMG fell as tumor size decreased, because diagnosis had to be made on the basis of morphologic criteria alone. With MRI, small tumors were diagnosed as well as large tumors, because diagnosis was based on a combination of morphologic criteria and the findings of the dynamic study. For a similar reason,
Vol. 198, No. 2, February 2004
Hata et al
Usefulness of MRI Mammography
195
Figure 6. Rate of sensitivity, specificity, and accuracy of detection of intraductal spread of breast cancer by means of magnetic resonance imaging (MRI), mammography (MMG), and ultrasonography (US). The sensitivity and accuracy of MRI was significantly superior to those of MMG and US, although the specificity of MRI was significantly inferior to MMG and US. Black bar, sensitivity; spotted bar, specificity; striped bar, accuracy. MRI versus MMG and US: sensitivity, p ⬍ 0.05; specificity, p ⬍ 0.05; accuracy, p ⬍ 0.05.
MRI could detect any histologic type of breast cancer, such as DCIS, IDC, ILC, and MC, although some cases of ductal carcinoma in situ or mucinous carcinoma were not detected on MMG or US. Ando and colleagues27 demonstrated that the pattern of tumor vascularity as shown by MRI also could help differentiate malignant tumors from benign lesions that were enhanced on contrast administration. In our study, breast surgeons who diagnosed breast cancer interpreted MRI images. Still, this is hardly considered to bias the results because we
Figure 7. Correlation between the extent of intraductal spread of breast cancer as measured by magnetic resonance imaging (MRI) and the pathologic measured distance. A significant correlation between the distance on MRI and pathologic distance was identified.
adopted highly objective criteria for the diagnosis of breast cancer. With respect to time-intensity curves in dynamic MRI of the breast, Kuhl and colleagues14 classified curve patterns as type I, which was a straight or curved time course with steady intensity increase; type II, a plateau of intensity with a sharp bend after the initial upstroke; or type III, a washout of intensity after the initial upstroke. According to their report, types I, II, and III were found in 83.0%, 11.5%, and 5.5% of benign solid tumors and fibrocystic changes, respectively, whereas types I, II, and III were found in 8.9%, 33.6%, and 57.4% of malignant lesions, respectively. Accordingly, in dynamic MRI of the breast, the pattern of the time-intensity curve is an important criterion in differentiating benign and malignant enhancing lesions. In our study, patterns of time intensity curve in dynamic MRI differed with histologic types. This finding is consistent with the fact that some malignant lesions, such as certain invasive ductal and lobular carcinomas and certain ductal carcinoma in situ lesions, are not enhanced rapidly, but their morphology nevertheless suggests the presence of malignancy.28,29 Kuhl and colleagues14 also found that lobular or scirrhous ductal invasive cancers could exhibit a benign pattern of time-intensity curve because of slow enhancement and poor angiogenic activity. About the diagnosis of breast cancer, we can summarize that MRI provides not only the morphologic diag-
196
Hata et al
Usefulness of MRI Mammography
nosis, but also the opportunity to evaluate vascularity with a contrast medium, which is impossible for other imaging modalities, such as MMG and US. With regard to intraductal spread of breast cancer, microscopic intraductal spread is often unrecognized during surgery and can cause local recurrence.2,3 Reported rates of local recurrence range from 5.5% to 27.9%.30-33 If local recurrence occurs after breastconserving surgery, the conserved breast is usually removed. Consequently, accurate assessment of the extent of intraductal spread is essential for determining the feasibility of breast-conserving surgery and the range of excision of the mammary gland. US and MMG have been used for evaluating the extent of breast cancer, but both have limitations. Realtime US must be performed by a highly skilled individual, and its interpretation is more complicated than that of MRI and MMG because fixed figures cannot be always shown. Satake and colleagues34 have reported that the sensitivity, specificity, and accuracy rates in the detection of intraductal spread with US were 89%, 76%, and 85%, respectively. In our study, the sensitivity rate of US for detecting intraductal spread was very low in comparison to MRI, probably because our criterion of intraductal spread on US was too simplified compared with that of Satake and colleagues.34 Given that MMG cannot identify intraductal spread that does not include minute calcifications, it was predictable that the sensitivity rate of MMG for detecting intraductal spread would be lower than MRI in our study. Satake and colleagues34 also reported that the sensitivity of MMG was lower than that of US and MRI. According to the recent literature, intraductal spread can be enhanced by contrast medium, thus making it detectable by dynamic MRI;35 the rate of the sensitivity ranges from 55% to 100%.4-9,10,35-38 Some authors have shown that MRI detects intraductal spread more accurately than MMG and US,10,20 as we indicated in our study. Recently, three-dimensional MRI has been used for detection of intraductal spread, and its usefulness beyond computed tomography or US has been reported.11,12 In 54 of 68 cases in which intraductal spread was detected by MRI and the extent was determined pathologically, a significant positive correlation between the distance on MRI and pathologic distance was identified. The distance of intraductal spread measured on MRI was greater than the actual pathologic extent in 38.9%
J Am Coll Surg
of cases. Fibrocystic disease around the intraductal spread and lymphocyte infiltration and fibrous proliferation associated with a tumor can lead to enhancement and overestimation of the distance on MRI, as other authors also have mentioned.6,19,36 Given that MRI overestimates the extent of intraductal spread, there is the potential risk of unnecessary mastectomies. This problem should be taken into account on the occasion of planning of breast-conserving operation with MRI. In conclusion, MRI can diagnose breast cancer as accurately as US and more accurately than MMG without respect to tumor size or histologic type. In addition, patterns of time-intensity curves may enable us to predict histologic type. MRI can detect intraductal spread of breast cancer more accurately than MMG and US. It is predictable that preoperative MRI reduces residual intraductal spread of breast cancer in the margin of conserved breast and the rate of local recurrence after breastconserving surgery. So, in view of the particular advantages of MRI in the diagnosis and preoperative characterization of breast cancer, MRI is considered indispensable for determining the appropriateness of breast-conserving surgery and for planning the surgical approach to prevent postoperative local recurrence. The clinical outcomes after operation planning with MRI of the breast remain unsolved. A further study is required to answer this question. Acknowledgment: We are deeply indebted to Akiko Tukahara, MD, at the Department of Radiology Hokkaido University Hospital, who consulted about the interpretation of magnetic resonance images.
REFERENCES 1. Heywang SH, Hahn D, Schmidt H, et al. MR imaging of the breast using gadolinium-DTPA. J Comput Assist Tomograr 1986;10:99–204. 2. Lindley R, Bulman A, Parsons P, et al. Histologic features predictive of an increased risk of early local recurrence after treatment of breast cancer by local tumor excision and radical radiotherapy. Surgery 1989;105:13–20. 3. Holland R, Connolly JL, Gelman R, et al. The presence of an extensive intraductal component following a limited excision correlates with prominent residual disease in the remainder of the breast. J Clin Oncol 1990;8:113–118. 4. Mumtaz H, Hall-Craggs MA, Davidson T, et al. Staging of symptomatic primary breast cancer with MR imaging. Am J Roentgenol 1997;169:417–424. 5. Hiramatu H, Enomoto K, Ikeda T, et al. The role of contrast-
Vol. 198, No. 2, February 2004
6.
7.
8.
9.
10.
11. 12. 13. 14.
15. 16. 17. 18. 19. 20. 21.
enhanced high resolution MRI in the surgical planning of breast cancer. Breast Cancer 1997;4:285–290. Date S. Diagnosis of intraductal spread of breast cancer by highresolution MR imaging: correlation between MR imaging and pathohistological findings. Nippon Igaku Hoshasen Gakkai Zasshi 1998;58:212–220. Drew PJ, Chatterjee S, Turnbull LW, et al. Dynamic contrast enhanced magnetic resonance imaging of the breast is superior to triple assessment for the pre-operative detection of multifocal breast cancer. Ann Surg Oncol 1999;6:599–603. Gatzemeier W, Liesch T, Stylianou A, et al. Preoperative MR mammography in breast carcinoma. Effect on operative treatment from the surgical viewpoint. Chirurg 1999;70:1460– 1468. Shiraishi A, Suzuki M, Nozu S, et al. Diagnosis of breast cancer extent and enhance patterns using 3D-dynamic imaging: correlation with intraductal component. Nippon Igaku Hoshasen Gakkai Zasshi 1999;59:122–130. Malur S, Wurdinger S, Moritz A, et al. Comparison of written reports of mammography, sonography and magnetic resonance mammography for preoperative evaluation of breast lesions, with special emphasis on magnetic resonance mammography. Breast Cancer Res 2001;3:55–60. Tamaki Y, Akashi-Tanaka S, Ishida T, et al. 3D imaging of intraductal spread of breast cancer and its clinical application for navigation surgery. Breast Cancer Res 2002;9:289–95. Nakahara H, Namba K, Wakamatsu H, et al. Extension of breast cancer: comparison of CT and MRI. Radiat Med 2002;20:17– 23. Hermanke LH, Sobin LH. International Union Against Cancer (UICC): TNM classification of malignant tumors, 4th ed. Berlin: Springer; 1989. Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR imaging. Are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 1989;211:101– 110. Heywang SH, Wolf A, Pruss E, et al. MR imaging of the breast with Gd-DTPA; use and limitations. Radiology 1989;171:95– 103. Kaiser WA, Zeitler E. MR imaging of the breast; Fast imaging sequences with and without Gd-DTPA. Preliminary observations. Radiology 1989;170:681–686. Stack JP, Redmond OM, Codd MB, et al. Breast disease: tissue characterization with Gd-DTPA enhancement profiles. Radiology 1990;174:491–494. WHO. Histological typing of breast tumours. International histological classification of tumours. 2nd ed. Geneva: World Health Organization; 1981. Gilles R, Guinebretiere JM, Lucidarme O, et al. Nonpalpable breast tumors: diagnosis with contrast-enhanced subtraction dynamic MR imaging. Radiology 1994;191:625–631. Boetes C, Mus RD, Holland R, et al. Breast tumor: comparative accuracy of MR imaging relative to mammography and US for demonstrating extent. Radiology 1995;197:743–747. Orel SG, Schnall MD. MR imaging of the breast for the detection, diagnosis, and staging of breast cancer. Radiology 2001; 220:13–30.
Hata et al
Usefulness of MRI Mammography
197
22. Bone B, Aspelin P, Bronge L, et al. Sensitivity and specificity of MR mammography with histopathological correlation in 250 breasts. Acta Radiol 1996;37:208–213. 23. Helbich TH, Becherer A, Trattnig S, et al. Differentiation of benign and malignant breast lesions: MR imaging versus Tc99m sestamibi scintimammography. Radiology 1997;202:421– 429. 24. Fischer U, Kopka L, Grabbe E. Breast carcinoma: effect of preoperative contrast-enhanced MR imaging on the therapeutic approach. Radiology 1999;213:881–888. 25. Wedegartner U, Bick U, Wortler K, et al. Differentiation between benign and malignant findings on MR-mammography: usefulness of morphological criteria. Eur Radiol 2001;11:1645– 1650. 26. Davis PL, Staiger MJ, Harris KB, et al. Breast cancer measurement with magnetic resonance imaging, ultrasonography, and mammography. Breast Cancer Res Treat 1996;37:1–9. 27. Ando Y, Fukatsu H, Ishiguchi T, et al. Diagnostic utility of tumor vascularity on magnetic resonance imaging of the breast. Magn Reson Imaging 2000;18:807–813. 28. Nunes LW, Schnall MD, Orel SG, et al. Breast MR imaging. Interpretation model. Radiology 1997;202:833–841. 29. Orel SG. Differentiating benign from malignant enhancing lesions identified at MR imaging of the breast: are time-signal intensity curves an accurate predictor? Radiology 1999;211: 5–7. 30. Abner AL, Recht A, Eberlein T, et al. Prognosis following salvage mastectomy for recurrence in the breast after conservative surgery and radiation therapy for early-stage breast cancer. J Clin Oncol 1993;11:44–48. 31. Fisher B, Bauer M, Margolese R, et al. Five-year results of a randomized clinical trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N Engl J Med 1985;14:312 665–673. 32. Lagios MD, Margolin FR, Westdahl PR, et al. Mammographically detected duct carcinoma in situ. Frequency of local recurrence following tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer 1989;63:15 618–624. 33. Fisher ER, Sass R, Fisher B, et al. Pathologic findings of the National Surgical Adjuvant Breast Project (Protocol 6). I. Intraductal carcinoma (DCIS). Cancer 1986;15(57):197–208. 34. Satake H, Shimamoto K, Sawaki A, et al. Role of ultrasonography in the detection of intraductal spread of breast cancer: correlation with pathologic findings, mammography and MR imaging. Eur Radiol 2000;10:1726–1732. 35. Gilles R, Zafrani B, Guinebretiere JM, et al. Ductal carcinoma in situ: MR imaging-histopathologic correlation. Radiology 1995;196:415–419. 36. Kawashima H. Fat suppressed dynamic MR imaging of breast cancer. Radiologic-pathologic correlation. Nippon Igaku Hoshasen Gakkai Zasshi 1996;56:459–569. 37. Orel SG, Mendonca MH, Reynolds C, et al. MR imaging of ductal carcinoma in situ. Radiology 1997;202:413–420. 38. Ando Y, Fukatsu H, Ishigaki T, et al. Intramammary extension of breast cancer: diagnosis with MR mammography. Breast Cancer 1998;5:291–300.