- Email: [email protected]
MR imaging of medullary carcinoma of the breast
European Journal of Radiology 70 (2009) 525–529
MR imaging of medullary carcinoma of the breast Junya Tominaga a,∗ , Hikaru Hama a , Noriko Kimura b , Shoki Takahashi c a
Department of Radiology, Tohoku Rohsai Hospital, 21-3-4 Dainohara Aoba-ku, Sendai 981-8563, Japan b Department of Pathology, Japan National Hospital Organization, Hakodate Hospital, 16-18 Kawahara-cho Hakodate, Hokkaido 041-8512, Japan c Department of Diagnostic Radiology, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan Received 8 October 2007; received in revised form 5 January 2008; accepted 28 January 2008
Abstract Purpose: To examine the magnetic resonance imaging (MRI) findings of medullary carcinoma of the breast and to correlate them with histopathologic features. Materials and methods: Eight patients were retrospectively evaluated with pathologically confirmed medullary carcinoma of the breast. T1weighted fat-saturated, T2-weighted fast spine echo, and gadolinium-enhanced fat-saturated fast spoiled gradient-echo images were obtained. Interpretation of the MRI findings was based on evaluation of the configuration, internal signal intensity, contrast enhancement, and type of the time–intensity curve. Results: Medullary carcinoma showed a lobular shape and a smooth margin, either homogenous or heterogeneous enhancement and delayed peripheral enhancement in the late phase on contrast-enhanced MRI, and either a plateau or washout type with rapid initial rise on the time–intensity curve of the dynamic study. Conclusion: Although the MRI findings showed a close relationship with histopathologic features of medullary carcinoma, it was difficult to differentiate medullary carcinoma from other histologic types of invasive breast carcinomas. © 2008 Elsevier Ireland Ltd. All rights reserved. Keywords: MRI; Medullary carcinoma; Delayed peripheral enhancement
1. Introduction Medullary carcinoma of the breast is an uncommon invasive breast cancer and previous studies reported that it accounts for less than 5% up to 7% of breast cancers [1]. Medullary carcinoma is defined by the World Health Organization (WHO) classification of breast tumors as “a well-circumscribed carcinoma composed of poorly differentiated cells arranged in large sheets, with no glandular structure, scant stroma and a prominent lymphoplasmacytic infiltrate”. It has a more favorable prognosis than infiltrating ductal carcinoma [2]. A series of 1490 patients managed with breast-conservation therapy that consisted of lumpectomy and radiation therapy at Yale University included 46 cases of medullary carcinoma. The 10-year distant relapse-free survival in the medullary cohort was significantly better than in the control group of infiltrating ductal carcinoma
∗
Corresponding author. Tel.: +81 22 717 7312; fax: +81 22 717 7316. E-mail address: [email protected] (J. Tominaga).
0720-048X/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2008.01.044
(94.9% vs. 77.5%, p = 0.028). This suggests that patients diagnosed with medullary carcinoma are appropriate candidates for conservative treatment with lumpectomy and radiation therapy and are associated with favorable prognosis in long-term outcome [3]. Therefore, obtaining a pre-operative diagnosis of medullary carcinoma would be a desirable goal. Several authors have reported on mammographic and ultrasonographic findings of medullary carcinoma of the breast, and concluded that it was not possible to differentiate medullary carcinoma from solid benign masses and other malignant lesions by those conventional diagnostic modalities [4,5]. Magnetic resonance imaging (MRI) may provide helpful information for differential diagnosis of breast tumors, but to our knowledge, characteristic features of medullary carcinoma has never been described in the MRI literature. In this retrospective study, the MRI findings of medullary carcinoma in eight patients were reviewed and compared with their histopathologic features. The aim of this study is to evaluate the potential of MRI to differentiate medullary carcinoma from the other types of breast cancer.
526
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
2. Materials and methods Breast MRI has been performed since October 2003 using our MR system, and eight patients were found with pathologically confirmed medullary carcinoma by analysis of specimens obtained by surgery between October 2003 and April 2007. They were all females and ranged in age from 39 to 56 years with a mean age of 48.9 years. MRI was performed after the fine needle aspiration cytology diagnosis of breast cancer in four patients to evaluate the tumor extent, while in the remaining four patients it was performed prior to aspiration cytology for the purpose of scrutinizing a new growing breast lump. MRI was performed using a 1.5-T MR scanner (Signa Twin Speed Excite 1.5 T, GE Medical Systems, Milwaukee, WI, USA). The affected side in each patient was examined using a commercially available breast coil with the patient in the prone position. MRI performed before the administration of contrast material comprised sagittal T1-weighted spinecho images (TR/TE, 640/11.8; thickness, 5 mm; gap, 1 mm; matrix, 256 × 224), and sagittal fat-saturated T2-weighted fast spin-echo images (4000/72.3; thickness, 5 mm; gap, 1 mm; matrix, 256 × 224). After the initial examination, dynamic contrast-enhanced images were obtained by using a fat-saturated fast spoiled gradient-echo sequence (150/2.6; flip angle, 60; thickness, 5 mm; gap, 1 mm; matrix, 256 × 224). A bolus of gadopentetate dimeglumine (Magnevist, Schering Berlin, Germany) was injected into the antecubital vein at a dose of 0.1 mmol/kg body weight within 10 s, followed by a 20-mL saline solution flush. Sequential multisection, whole-breast images were obtained in the sagittal plane at 20 s intervals for 160 s. For analysis of enhancement kinetics, time–intensity curves were plotted from the signal intensity values obtained in the lesion including the most enhanced part. Dynamic contrastenhanced images at 40 s after injection of contrast material were evaluated as the contrast-enhanced MRI at the early phase. In all patients, sagittal fat-saturated three-dimensional fast spoiled gradient-echo images (19.34/4.19; flip angle, 35; thickness, 2 mm; gap, 1 mm; matrix, 384 × 384) were obtained at 8 min after injection of contrast material as the contrast-enhanced
MRI at the late phase. The field of view was 18 cm in all sequences. Interpretation of the MRI findings was based on evaluation of the configuration, internal signal intensity, contrast enhancement and type of the time–intensity curve. All of the tumors were surgically resected and subjected to routine histopathologic examination with hematoxylin–eosin (H–E) staining. The MRI and histopathologic findings were studied for correlations. 3. Results The MRI and pathologic findings of medullary carcinoma of the breast in the eight patients are shown in Table 1. All of the tumors were iso-intense on T1-weighted and iso-intense or slightly hyper-intense on fat-saturated T2-weighted images compared with the surrounding non-tumor breast tissue. However, all tumors clearly exhibited lesions with lobular shapes and smooth margins upon contrast enhancement. The maximum size of the tumors ranged from 13 to 36 mm. Four tumors (two tumors before aspiration cytology and the other two after aspiration cytology; patients 1–4) showed homogenous enhancement, and the remaining four (two tumors before aspiration cytology and the other two after aspiration cytology; patients 5–8) showed heterogeneous enhancement with small non-enhanced spots on contrast-enhanced MRI. All eight tumors had delayed peripheral enhancement in the late phase on contrast-enhanced MRI. The time–intensity curve of the dynamic study showed either a plateau type with rapid initial rise (rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a plateau) or a washout type with rapid initial rise (rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a decrease). On histopathologic examination by analysis of specimens obtained by surgery, all eight tumors were circumscribed by the surrounding compressed fibrous tissue with prominent lymphocytic infiltration. Four tumors (patients 1–4) had uniform distribution of cancer cells without necrosis, whereas the other four tumors (patients 5–8) had necrotic areas inside lesions with cystic foci. All of the first four tumors showed
Table 1 MRI findings and pathologic features of eight patients with medullary carcinoma of the breast Patient
MRI findings
Pathologic findings
No.
Age (years)
Maximum size (mm)
Configuration
Contrast enhancement
Late phase
TIC
Necrosis/cystic foci
Lymphocytic infiltrate in surrounding tissue
1 2 3 4 5 6 7 8
45 39 49 51 50 56 45 56
13 32 19 14 32 24 17 36
Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth
Homo Homo Homo Homo Hetero with NES Hetero with NES Hetero with NES Hetero with NES
DPE(+) DPE(+) DPE(+) DPE(+) DPE(+) DPE(+) DPE(+) DPE(+)
Plateau Washout Washout Plateau Washout Washout Plateau Washout
(−) (−) (−) (−) (+) (+) (+) (+)
(+) (+) (+) (+) (+) (+) (+) (+)
Lobular/smooth = a lobular shape and a smooth margin. Homo = homogenous enhancement; Hetero = heterogeneous enhancement; NES = non-enhanced spot; DPE = delayed peripheral enhancement. TIC = type of the time–intensity curve. Plateau = plateau type: rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a plateau. Washout = washout type: rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a decrease.
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
527
Fig. 1. Medullary carcinomas in a 45-year-old female (patient 1). (A) Sagittal contrast-enhanced MR image in the early phase at 40 s after administration of contrast material shows homogeneous enhancement. (B) Sagittal contrast-enhanced MR image in the late phase at 8 min after administration of contrast material shows a lobular shape and a smooth margin with delayed peripheral enhancement, which may represent relative clearance of enhancement in the central portion of the tumor. (C) Photomicrograph of the marginal area of the tumor. The tumor is composed of poorly differentiated cells with pleomorphic nuclei, prominent nucleoli with typical syncytial growth in which the cells are arranged in broad sheets. Cancer cells are uniformly distributed in each sheet without necrotic areas. The tumor is well circumscribed by the surrounding compressed tissue with prominent lymphocytic infiltrate on the right side of the view. The border of the tumor is shown by arrows. H–E stain, ×100.
homogeneous enhancement and the other four tumors showed heterogeneous enhancement with small non-enhanced spots on contrast-enhanced MRI. Fig. 1 shows the MR images before aspiration cytology and a microscopic view of the tumor of patient 1. The tumor was located in the upper right outer quadrant, and exhibited a lesion with a lobular shape and a smooth margin. On contrastenhanced MRI, the tumor showed homogeneous enhancement in the early phase (Fig. 1A) and delayed peripheral enhancement in the late phase (Fig. 1B) on contrast-enhanced MRI. Histologically, the tumor was composed of poorly differentiated cancer cells arranged in large sheets, and it was well circumscribed by the surrounding compressed tissue with prominent lymphocytic infiltrates. Necrosis was not observed (Fig. 1C). Fig. 2 shows the MR images before aspiration cytology and a microscopic view of the tumor of patient 5. The tumor was
located in the upper left outer quadrant. It exhibited a lesion with a lobular shape and a smooth margin, and had heterogeneous enhancement with a small non-enhanced spots in the early phase (Fig. 2A) and delayed peripheral enhancement in the late phase (Fig. 2B) on contrast-enhanced MRI. Histologically, the tumor was composed of poorly differentiated cancer cells with prominent lymphocytic infiltrates. Necrotic areas and cystic foci were observed in the tumor (Fig. 2C). 4. Discussion Although medullary carcinoma has a unique morphology among breast cancers, it is not possible to differentiate medullary carcinoma from solid benign masses and other malignant lesions by conventional mammography and ultrasonography [4,5]. Considering the possibility that MRI may provide a clue to the
528
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
Fig. 2. Medullary carcinomas in a 50-year-old female (patient 5). (A) Sagittal contrast-enhanced MR image in the early phase at 40 s after administration of contrast material shows heterogeneous enhancement with non-enhanced spots. (B) Sagittal contrast-enhanced MR image in the late phase at 8 min after administration of contrast material shows a lobular shape and a smooth margin with delayed peripheral enhancement. (C) Photomicrograph of the central area of the tumor. The tumor is composed of poorly differentiated cells with pleomorphic nuclei, prominent nucleoli with typical syncytial growth in which the cells are arranged in broad sheets. Necrotic areas are evident in the center of the view (asterisk). H–E stain, ×100.
diagnosis of medullary carcinoma, we retrospectively studied the MRI findings of eight patients with medullary carcinoma and searched for correlations with their histopathologic features. On contrast-enhanced MRI in the present study, all of the tumors clearly showed a lesion with a lobular shape and a smooth margin, which presumably reflected the pathologic features of expansive growth with a border circumscribed by the surrounding compressed tissue. Spiculated borders were not present, probably because the tumors did not pathologically infiltrate to the surrounding tissue with desmoplastic reaction. The tumors with homogeneous enhancement on contrast-enhanced MRI corresponded to uniform distribution of cancer cells without necrotic areas inside the tumors, while the tumors with heterogeneous enhancement with non-enhanced spots corresponded to necrotic and cystic foci inside the tumors. In the two patients who underwent MRI after aspiration cytology, we admit that the heterogeneous enhancement with non-enhanced spots on MRI and histopathological tumor necrosis might have been
caused by the procedure. However, necrosis is not uncommon in medullary carcinomas, and as the extent of necrosis increases, there is a greater likelihood that the tumor will develop cystic foci [1], which by MRI will appear with heterogeneous enhancement with non-enhanced spots. Many more cases with imaging-pathologic correlation may be needed to draw this conclusion. In the late phase on contrast-enhanced MRI, all of the medullary carcinomas in this study showed delayed peripheral enhancement. Previous reports also described this type of enhancement in breast cancer, and suggested that it represented the presence of a high degree of surrounding inflammatory changes, the rim of the compressed connective tissue and expansive growth pattern [6,7]. With lymphocytic reactions, the edges of the medullary tumors in the present study showed smooth and rounded margins that appeared to push aside rather than infiltrating the adjacent mammary parenchyma, as has been described in the literature [1]. Such pathological features might
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
be attributable to the delayed peripheral enhancement found in common in our cases. However, this pattern of enhancement is not specific to medullary carcinoma because it has been seen in the other histologic types of breast cancers as well as in benign lesions [6,7]. All the tumors in this study showed rapid initial rise on the time–intensity curve (either a plateau or washout type) in the dynamic study. This pattern of enhancement correlated with higher microvascular counts on immunohistochemical staining with CD34 antigen, which were recorded for malignant lesions rather than for benign lesions [8]. Hence, evaluation for enhancement kinetics by time–intensity curve on MRI would be helpful for differentiation of medullary carcinoma as well as other subtypes of breast cancer from benign lesions. Invasive ductal carcinoma, not otherwise specified (NOS), is the largest group of malignant breast tumors and constitutes 65–80% of breast carcinomas [1]. The majority of these have irregular, ill-defined or spiculated margins, while 10–16% have lobulated margins like medullary carcinoma [8–10]. Moreover, invasive ductal carcinomas also show either a plateau or washout type with rapid initial rise on the time–intensity curve [11,12], and approximately half of them exhibited delayed peripheral enhancement [7]. Regrettably therefore, MRI features found in our cases cannot be regarded as specific to medullary carcinoma but may well be seen in other types of breast cancer including invasive ductal carcinoma (NOS). In this respect, a special subtype of invasive ductal carcinoma (NOS), “infiltrating ductal carcinoma with medullary features” may be worth mentioning. This subtype of tumor shares a common histopathological feature with medullary carcinoma in showing a clearly circumscribed interface with the surrounding compressed tissue with lymphocytic infiltrates, although it can be differentiated from medullary carcinoma by its microscopic structures at cellular level. The MRI findings of such cases can be very similar to those of medullary carcinoma making their differentiation even difficult. Admittedly, the number of patients available for this study was small. A strict correlation between MRI findings and histopathologic features was difficult, because each of the sagittal images on MRI did not always match the plane of section of the pathologic specimen. Nonetheless, it is felt that this exploratory study has shown a promising potential for MRI as a pre-operative diagnostic tool for differentiating different types of breast carcinomas. Therefore, further study is anticipated with a larger number of patients combined with a stricter correlation between MRI and histopathology.
529
In conclusion, medullary carcinoma of the breast showed a lobular shape and a smooth margin, either homogenous or heterogeneous contrast enhancement, delayed peripheral enhancement in the late phase on contrast-enhanced MRI, and rapid initial rise (either plateau or washout type) on the time–intensity curve of the dynamic study. Although such MRI findings appeared closely related to pathologic features suggestive of medullary carcinoma, its differentiation from some other types of invasive breast cancers remains difficult by MRI alone. References [1] Medullary carcinoma. In: Rosen PP, editor. Rosen’s breast pathology. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001. p. 405– 424. [2] Reinfuss M, Stelmach A, Mitus J, Rys J, Duda K. Typical medullary carcinoma of the breast: a clinical and pathological analysis of 52 cases. J Surg Oncol 1995;60:89–94. [3] Vu-Nishino H, Tavassoli FA, Ahrens WA, Haffty BG. Clinicopathologic features and long-term outcome of patients with medullary breast carcinoma managed with breast-conserving therapy (BCT). Int J Radiat Oncol Biol Phys 2005;62:1040–7. [4] Meyer JE, Amin E, Lindfors KK, Lipman JC, Stomper PC, Genest D. Medullary carcinoma of the breast: mammographic and US appearance. Radiology 1989;170:79–82. [5] Liberman L, LaTrenta LR, Samli B, Morris EA, Abramson AF, Dershaw DD. Overdiagnosis of medullary carcinoma: a mammographic–pathologic correlative study. Radiology 1996;201:443–6. [6] Matsubayashi R, Matsuo Y, Edakuni G, Satoh T, Tokunaga O, Kudo S. Breast masses with peripheral rim enhancement on dynamic contrastenhanced MR images: correlation of MR findings with histologic features and expression of growth factors. Radiology 2000;217:841– 8. [7] Buadu LD, Murakami J, Murayama S, et al. Patterns of peripheral enhancement in breast masses: correlation of findings on contrast medium enhanced MRI with histologic features and tumor angiogenesis. J Comput Assist Tomogr 1997;21:421–30. [8] Buadu LD, Murakami J, Murayama S, et al. Breast lesions: correlation of contrast medium enhancement pattern on MR images with histopathologic findings and tumor angiogenesis. Radiology 1996;200:639– 49. [9] Nunes LW, Schnall MD, Orel SG, et al. Correlation of lesion appearance and histologic findings for the nodes a breast MR imaging interpretation model. Radiographics 1999;19:79–92. [10] Takeda Y, Yoshikawa K. Contrast-enhanced dynamic MR imaging parameters and histological types of invasive ductal carcinoma of breast. Biomed Pharmacother 2005;59:115–21. [11] Fobben ES, Rubin CZ, Kalisher L, Dembner AG, Seltzer MH, Santoro EJ. Breast MR imaging with commercially available techniques: radiologic–pathologic correlatoin. Radiology 1995;196:143– 52. [12] Kusama R, Takayama F, Tsuchiya S. MRI of the breast: comparison of MRI signals and histological characteristics of the same slices. Med Mol Morphol 2005;38:204–15.
MR imaging of medullary carcinoma of the breast Junya Tominaga a,∗ , Hikaru Hama a , Noriko Kimura b , Shoki Takahashi c a
Department of Radiology, Tohoku Rohsai Hospital, 21-3-4 Dainohara Aoba-ku, Sendai 981-8563, Japan b Department of Pathology, Japan National Hospital Organization, Hakodate Hospital, 16-18 Kawahara-cho Hakodate, Hokkaido 041-8512, Japan c Department of Diagnostic Radiology, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan Received 8 October 2007; received in revised form 5 January 2008; accepted 28 January 2008
Abstract Purpose: To examine the magnetic resonance imaging (MRI) findings of medullary carcinoma of the breast and to correlate them with histopathologic features. Materials and methods: Eight patients were retrospectively evaluated with pathologically confirmed medullary carcinoma of the breast. T1weighted fat-saturated, T2-weighted fast spine echo, and gadolinium-enhanced fat-saturated fast spoiled gradient-echo images were obtained. Interpretation of the MRI findings was based on evaluation of the configuration, internal signal intensity, contrast enhancement, and type of the time–intensity curve. Results: Medullary carcinoma showed a lobular shape and a smooth margin, either homogenous or heterogeneous enhancement and delayed peripheral enhancement in the late phase on contrast-enhanced MRI, and either a plateau or washout type with rapid initial rise on the time–intensity curve of the dynamic study. Conclusion: Although the MRI findings showed a close relationship with histopathologic features of medullary carcinoma, it was difficult to differentiate medullary carcinoma from other histologic types of invasive breast carcinomas. © 2008 Elsevier Ireland Ltd. All rights reserved. Keywords: MRI; Medullary carcinoma; Delayed peripheral enhancement
1. Introduction Medullary carcinoma of the breast is an uncommon invasive breast cancer and previous studies reported that it accounts for less than 5% up to 7% of breast cancers [1]. Medullary carcinoma is defined by the World Health Organization (WHO) classification of breast tumors as “a well-circumscribed carcinoma composed of poorly differentiated cells arranged in large sheets, with no glandular structure, scant stroma and a prominent lymphoplasmacytic infiltrate”. It has a more favorable prognosis than infiltrating ductal carcinoma [2]. A series of 1490 patients managed with breast-conservation therapy that consisted of lumpectomy and radiation therapy at Yale University included 46 cases of medullary carcinoma. The 10-year distant relapse-free survival in the medullary cohort was significantly better than in the control group of infiltrating ductal carcinoma
∗
Corresponding author. Tel.: +81 22 717 7312; fax: +81 22 717 7316. E-mail address: [email protected] (J. Tominaga).
0720-048X/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2008.01.044
(94.9% vs. 77.5%, p = 0.028). This suggests that patients diagnosed with medullary carcinoma are appropriate candidates for conservative treatment with lumpectomy and radiation therapy and are associated with favorable prognosis in long-term outcome [3]. Therefore, obtaining a pre-operative diagnosis of medullary carcinoma would be a desirable goal. Several authors have reported on mammographic and ultrasonographic findings of medullary carcinoma of the breast, and concluded that it was not possible to differentiate medullary carcinoma from solid benign masses and other malignant lesions by those conventional diagnostic modalities [4,5]. Magnetic resonance imaging (MRI) may provide helpful information for differential diagnosis of breast tumors, but to our knowledge, characteristic features of medullary carcinoma has never been described in the MRI literature. In this retrospective study, the MRI findings of medullary carcinoma in eight patients were reviewed and compared with their histopathologic features. The aim of this study is to evaluate the potential of MRI to differentiate medullary carcinoma from the other types of breast cancer.
526
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
2. Materials and methods Breast MRI has been performed since October 2003 using our MR system, and eight patients were found with pathologically confirmed medullary carcinoma by analysis of specimens obtained by surgery between October 2003 and April 2007. They were all females and ranged in age from 39 to 56 years with a mean age of 48.9 years. MRI was performed after the fine needle aspiration cytology diagnosis of breast cancer in four patients to evaluate the tumor extent, while in the remaining four patients it was performed prior to aspiration cytology for the purpose of scrutinizing a new growing breast lump. MRI was performed using a 1.5-T MR scanner (Signa Twin Speed Excite 1.5 T, GE Medical Systems, Milwaukee, WI, USA). The affected side in each patient was examined using a commercially available breast coil with the patient in the prone position. MRI performed before the administration of contrast material comprised sagittal T1-weighted spinecho images (TR/TE, 640/11.8; thickness, 5 mm; gap, 1 mm; matrix, 256 × 224), and sagittal fat-saturated T2-weighted fast spin-echo images (4000/72.3; thickness, 5 mm; gap, 1 mm; matrix, 256 × 224). After the initial examination, dynamic contrast-enhanced images were obtained by using a fat-saturated fast spoiled gradient-echo sequence (150/2.6; flip angle, 60; thickness, 5 mm; gap, 1 mm; matrix, 256 × 224). A bolus of gadopentetate dimeglumine (Magnevist, Schering Berlin, Germany) was injected into the antecubital vein at a dose of 0.1 mmol/kg body weight within 10 s, followed by a 20-mL saline solution flush. Sequential multisection, whole-breast images were obtained in the sagittal plane at 20 s intervals for 160 s. For analysis of enhancement kinetics, time–intensity curves were plotted from the signal intensity values obtained in the lesion including the most enhanced part. Dynamic contrastenhanced images at 40 s after injection of contrast material were evaluated as the contrast-enhanced MRI at the early phase. In all patients, sagittal fat-saturated three-dimensional fast spoiled gradient-echo images (19.34/4.19; flip angle, 35; thickness, 2 mm; gap, 1 mm; matrix, 384 × 384) were obtained at 8 min after injection of contrast material as the contrast-enhanced
MRI at the late phase. The field of view was 18 cm in all sequences. Interpretation of the MRI findings was based on evaluation of the configuration, internal signal intensity, contrast enhancement and type of the time–intensity curve. All of the tumors were surgically resected and subjected to routine histopathologic examination with hematoxylin–eosin (H–E) staining. The MRI and histopathologic findings were studied for correlations. 3. Results The MRI and pathologic findings of medullary carcinoma of the breast in the eight patients are shown in Table 1. All of the tumors were iso-intense on T1-weighted and iso-intense or slightly hyper-intense on fat-saturated T2-weighted images compared with the surrounding non-tumor breast tissue. However, all tumors clearly exhibited lesions with lobular shapes and smooth margins upon contrast enhancement. The maximum size of the tumors ranged from 13 to 36 mm. Four tumors (two tumors before aspiration cytology and the other two after aspiration cytology; patients 1–4) showed homogenous enhancement, and the remaining four (two tumors before aspiration cytology and the other two after aspiration cytology; patients 5–8) showed heterogeneous enhancement with small non-enhanced spots on contrast-enhanced MRI. All eight tumors had delayed peripheral enhancement in the late phase on contrast-enhanced MRI. The time–intensity curve of the dynamic study showed either a plateau type with rapid initial rise (rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a plateau) or a washout type with rapid initial rise (rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a decrease). On histopathologic examination by analysis of specimens obtained by surgery, all eight tumors were circumscribed by the surrounding compressed fibrous tissue with prominent lymphocytic infiltration. Four tumors (patients 1–4) had uniform distribution of cancer cells without necrosis, whereas the other four tumors (patients 5–8) had necrotic areas inside lesions with cystic foci. All of the first four tumors showed
Table 1 MRI findings and pathologic features of eight patients with medullary carcinoma of the breast Patient
MRI findings
Pathologic findings
No.
Age (years)
Maximum size (mm)
Configuration
Contrast enhancement
Late phase
TIC
Necrosis/cystic foci
Lymphocytic infiltrate in surrounding tissue
1 2 3 4 5 6 7 8
45 39 49 51 50 56 45 56
13 32 19 14 32 24 17 36
Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth Lobular/smooth
Homo Homo Homo Homo Hetero with NES Hetero with NES Hetero with NES Hetero with NES
DPE(+) DPE(+) DPE(+) DPE(+) DPE(+) DPE(+) DPE(+) DPE(+)
Plateau Washout Washout Plateau Washout Washout Plateau Washout
(−) (−) (−) (−) (+) (+) (+) (+)
(+) (+) (+) (+) (+) (+) (+) (+)
Lobular/smooth = a lobular shape and a smooth margin. Homo = homogenous enhancement; Hetero = heterogeneous enhancement; NES = non-enhanced spot; DPE = delayed peripheral enhancement. TIC = type of the time–intensity curve. Plateau = plateau type: rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a plateau. Washout = washout type: rapid and strong enhancement was seen within the first 2 min after injection of contrast material, followed by a decrease.
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
527
Fig. 1. Medullary carcinomas in a 45-year-old female (patient 1). (A) Sagittal contrast-enhanced MR image in the early phase at 40 s after administration of contrast material shows homogeneous enhancement. (B) Sagittal contrast-enhanced MR image in the late phase at 8 min after administration of contrast material shows a lobular shape and a smooth margin with delayed peripheral enhancement, which may represent relative clearance of enhancement in the central portion of the tumor. (C) Photomicrograph of the marginal area of the tumor. The tumor is composed of poorly differentiated cells with pleomorphic nuclei, prominent nucleoli with typical syncytial growth in which the cells are arranged in broad sheets. Cancer cells are uniformly distributed in each sheet without necrotic areas. The tumor is well circumscribed by the surrounding compressed tissue with prominent lymphocytic infiltrate on the right side of the view. The border of the tumor is shown by arrows. H–E stain, ×100.
homogeneous enhancement and the other four tumors showed heterogeneous enhancement with small non-enhanced spots on contrast-enhanced MRI. Fig. 1 shows the MR images before aspiration cytology and a microscopic view of the tumor of patient 1. The tumor was located in the upper right outer quadrant, and exhibited a lesion with a lobular shape and a smooth margin. On contrastenhanced MRI, the tumor showed homogeneous enhancement in the early phase (Fig. 1A) and delayed peripheral enhancement in the late phase (Fig. 1B) on contrast-enhanced MRI. Histologically, the tumor was composed of poorly differentiated cancer cells arranged in large sheets, and it was well circumscribed by the surrounding compressed tissue with prominent lymphocytic infiltrates. Necrosis was not observed (Fig. 1C). Fig. 2 shows the MR images before aspiration cytology and a microscopic view of the tumor of patient 5. The tumor was
located in the upper left outer quadrant. It exhibited a lesion with a lobular shape and a smooth margin, and had heterogeneous enhancement with a small non-enhanced spots in the early phase (Fig. 2A) and delayed peripheral enhancement in the late phase (Fig. 2B) on contrast-enhanced MRI. Histologically, the tumor was composed of poorly differentiated cancer cells with prominent lymphocytic infiltrates. Necrotic areas and cystic foci were observed in the tumor (Fig. 2C). 4. Discussion Although medullary carcinoma has a unique morphology among breast cancers, it is not possible to differentiate medullary carcinoma from solid benign masses and other malignant lesions by conventional mammography and ultrasonography [4,5]. Considering the possibility that MRI may provide a clue to the
528
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
Fig. 2. Medullary carcinomas in a 50-year-old female (patient 5). (A) Sagittal contrast-enhanced MR image in the early phase at 40 s after administration of contrast material shows heterogeneous enhancement with non-enhanced spots. (B) Sagittal contrast-enhanced MR image in the late phase at 8 min after administration of contrast material shows a lobular shape and a smooth margin with delayed peripheral enhancement. (C) Photomicrograph of the central area of the tumor. The tumor is composed of poorly differentiated cells with pleomorphic nuclei, prominent nucleoli with typical syncytial growth in which the cells are arranged in broad sheets. Necrotic areas are evident in the center of the view (asterisk). H–E stain, ×100.
diagnosis of medullary carcinoma, we retrospectively studied the MRI findings of eight patients with medullary carcinoma and searched for correlations with their histopathologic features. On contrast-enhanced MRI in the present study, all of the tumors clearly showed a lesion with a lobular shape and a smooth margin, which presumably reflected the pathologic features of expansive growth with a border circumscribed by the surrounding compressed tissue. Spiculated borders were not present, probably because the tumors did not pathologically infiltrate to the surrounding tissue with desmoplastic reaction. The tumors with homogeneous enhancement on contrast-enhanced MRI corresponded to uniform distribution of cancer cells without necrotic areas inside the tumors, while the tumors with heterogeneous enhancement with non-enhanced spots corresponded to necrotic and cystic foci inside the tumors. In the two patients who underwent MRI after aspiration cytology, we admit that the heterogeneous enhancement with non-enhanced spots on MRI and histopathological tumor necrosis might have been
caused by the procedure. However, necrosis is not uncommon in medullary carcinomas, and as the extent of necrosis increases, there is a greater likelihood that the tumor will develop cystic foci [1], which by MRI will appear with heterogeneous enhancement with non-enhanced spots. Many more cases with imaging-pathologic correlation may be needed to draw this conclusion. In the late phase on contrast-enhanced MRI, all of the medullary carcinomas in this study showed delayed peripheral enhancement. Previous reports also described this type of enhancement in breast cancer, and suggested that it represented the presence of a high degree of surrounding inflammatory changes, the rim of the compressed connective tissue and expansive growth pattern [6,7]. With lymphocytic reactions, the edges of the medullary tumors in the present study showed smooth and rounded margins that appeared to push aside rather than infiltrating the adjacent mammary parenchyma, as has been described in the literature [1]. Such pathological features might
J. Tominaga et al. / European Journal of Radiology 70 (2009) 525–529
be attributable to the delayed peripheral enhancement found in common in our cases. However, this pattern of enhancement is not specific to medullary carcinoma because it has been seen in the other histologic types of breast cancers as well as in benign lesions [6,7]. All the tumors in this study showed rapid initial rise on the time–intensity curve (either a plateau or washout type) in the dynamic study. This pattern of enhancement correlated with higher microvascular counts on immunohistochemical staining with CD34 antigen, which were recorded for malignant lesions rather than for benign lesions [8]. Hence, evaluation for enhancement kinetics by time–intensity curve on MRI would be helpful for differentiation of medullary carcinoma as well as other subtypes of breast cancer from benign lesions. Invasive ductal carcinoma, not otherwise specified (NOS), is the largest group of malignant breast tumors and constitutes 65–80% of breast carcinomas [1]. The majority of these have irregular, ill-defined or spiculated margins, while 10–16% have lobulated margins like medullary carcinoma [8–10]. Moreover, invasive ductal carcinomas also show either a plateau or washout type with rapid initial rise on the time–intensity curve [11,12], and approximately half of them exhibited delayed peripheral enhancement [7]. Regrettably therefore, MRI features found in our cases cannot be regarded as specific to medullary carcinoma but may well be seen in other types of breast cancer including invasive ductal carcinoma (NOS). In this respect, a special subtype of invasive ductal carcinoma (NOS), “infiltrating ductal carcinoma with medullary features” may be worth mentioning. This subtype of tumor shares a common histopathological feature with medullary carcinoma in showing a clearly circumscribed interface with the surrounding compressed tissue with lymphocytic infiltrates, although it can be differentiated from medullary carcinoma by its microscopic structures at cellular level. The MRI findings of such cases can be very similar to those of medullary carcinoma making their differentiation even difficult. Admittedly, the number of patients available for this study was small. A strict correlation between MRI findings and histopathologic features was difficult, because each of the sagittal images on MRI did not always match the plane of section of the pathologic specimen. Nonetheless, it is felt that this exploratory study has shown a promising potential for MRI as a pre-operative diagnostic tool for differentiating different types of breast carcinomas. Therefore, further study is anticipated with a larger number of patients combined with a stricter correlation between MRI and histopathology.
529
In conclusion, medullary carcinoma of the breast showed a lobular shape and a smooth margin, either homogenous or heterogeneous contrast enhancement, delayed peripheral enhancement in the late phase on contrast-enhanced MRI, and rapid initial rise (either plateau or washout type) on the time–intensity curve of the dynamic study. Although such MRI findings appeared closely related to pathologic features suggestive of medullary carcinoma, its differentiation from some other types of invasive breast cancers remains difficult by MRI alone. References [1] Medullary carcinoma. In: Rosen PP, editor. Rosen’s breast pathology. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001. p. 405– 424. [2] Reinfuss M, Stelmach A, Mitus J, Rys J, Duda K. Typical medullary carcinoma of the breast: a clinical and pathological analysis of 52 cases. J Surg Oncol 1995;60:89–94. [3] Vu-Nishino H, Tavassoli FA, Ahrens WA, Haffty BG. Clinicopathologic features and long-term outcome of patients with medullary breast carcinoma managed with breast-conserving therapy (BCT). Int J Radiat Oncol Biol Phys 2005;62:1040–7. [4] Meyer JE, Amin E, Lindfors KK, Lipman JC, Stomper PC, Genest D. Medullary carcinoma of the breast: mammographic and US appearance. Radiology 1989;170:79–82. [5] Liberman L, LaTrenta LR, Samli B, Morris EA, Abramson AF, Dershaw DD. Overdiagnosis of medullary carcinoma: a mammographic–pathologic correlative study. Radiology 1996;201:443–6. [6] Matsubayashi R, Matsuo Y, Edakuni G, Satoh T, Tokunaga O, Kudo S. Breast masses with peripheral rim enhancement on dynamic contrastenhanced MR images: correlation of MR findings with histologic features and expression of growth factors. Radiology 2000;217:841– 8. [7] Buadu LD, Murakami J, Murayama S, et al. Patterns of peripheral enhancement in breast masses: correlation of findings on contrast medium enhanced MRI with histologic features and tumor angiogenesis. J Comput Assist Tomogr 1997;21:421–30. [8] Buadu LD, Murakami J, Murayama S, et al. Breast lesions: correlation of contrast medium enhancement pattern on MR images with histopathologic findings and tumor angiogenesis. Radiology 1996;200:639– 49. [9] Nunes LW, Schnall MD, Orel SG, et al. Correlation of lesion appearance and histologic findings for the nodes a breast MR imaging interpretation model. Radiographics 1999;19:79–92. [10] Takeda Y, Yoshikawa K. Contrast-enhanced dynamic MR imaging parameters and histological types of invasive ductal carcinoma of breast. Biomed Pharmacother 2005;59:115–21. [11] Fobben ES, Rubin CZ, Kalisher L, Dembner AG, Seltzer MH, Santoro EJ. Breast MR imaging with commercially available techniques: radiologic–pathologic correlatoin. Radiology 1995;196:143– 52. [12] Kusama R, Takayama F, Tsuchiya S. MRI of the breast: comparison of MRI signals and histological characteristics of the same slices. Med Mol Morphol 2005;38:204–15.