- Email: [email protected]
Diagnosis of breast cancer at dynamic MRI in patients with breast augmentation by paraffin or silicone injection
Clinical Radiology (2009) 64, 1175e1180
ORIGINAL PAPER
Diagnosis of breast cancer at dynamic MRI in patients with breast augmentation by paraffin or silicone injection J.H. Youk, E.J. Son*, E.-K. Kim, J.-A. Kim, M.J. Kim, J.Y. Kwak, S.M. Lee Department of Radiology, Research Institute of Radiological Science, Yonsei University College of Medicine, Seoul, Republic of Korea Received 10 February 2009; received in revised form 30 April 2009; accepted 11 May 2009
AIM: To determine the diagnostic performance of dynamic magnetic resonance imaging (MRI) for breast cancer in breasts augmented with liquid paraffin or silicone injection. MATERIALS AND METHODS: Among 62 patients with breast augmentation by liquid paraffin or silicone injection who had undergone dynamic breast MRI at our institution, 27 women, who had pathological diagnosis or at least 1-year MRI follow-up, were included in this retrospective study and their MRI images were reviewed. For enhancing lesions on MRI, the morphological features, enhancement kinetics, and BI-RADS assessment category were analysed. The lesion characteristics at MRI were correlated with the final diagnosis based on the histopathological result or at least 1-year MRI follow-up. RESULTS: Of the 27 patients, 17 enhancing lesions in 13 patients were found on MRI. All six lesions that were confirmed as malignancy showed suspicious morphological findings and type 2 or 3 enhancement kinetics, assigned to BIRADS category 4 or 5. Of the remaining 11 benign lesions, 10 showed benign-favouring morphological findings, and all showed type 1 enhancement kinetics, assigned to BI-RADS category 2 or 4. CONCLUSION: In patients with breasts injected with foreign material, MRI was used to successfully diagnose malignant breast lesions and could be the diagnostic method of choice. Analysis of the morphological and kinetic features at MRI in conjunction with clinical findings is essential. ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction The direct injection of foreign material, such as liquid paraffin and liquid silicone, into the breast was first used for breast augmentation in the early 1900s for paraffin and the 1960s for silicone.1 However, it has been largely abandoned owing to serious complications associated with the procedure, including painful nodularity, skin sloughing, migration of materials, granulomatous reactions, embolism, and even death.2 Nevertheless, this procedure has been performed illicitly by non-physicians and has appealed * Guarantor and correspondent: E. J. Son, Department of Radiology, Gangnam Severance Hospital, 146-92, Dogok-Dong, Kangnam-Ku, Seoul 135-720, Republic of Korea. Tel.: þ82 2 2019 3510; fax: þ82 2 3462 5472. E-mail address: [email protected] (E.J. Son).
to those who were not knowledgeable about its risks and complications because it is cheap, quick, and simple to perform.3,4 In the breast, the injected foreign materials cause a foreign-body granulomatous reaction and fibrosis with hard, nodular breast masses and architectural distortion mimicking neoplasm.5,6 Therefore, the role of conventional methods, such as physical, mammographic, and sonographic examination, is limited for the diagnosis of breast cancer in patients with breasts augmented by foreign materials, which causes a delay in the diagnosis.2 Dynamic contrast-enhanced magnetic resonance imaging (MRI) is thought to be an appropriate method for evaluation of lesions in such injected breasts, by virtue of its high temporal and spatial resolution.2,4,5,7,8 Although MRI features of breast paraffinoma or silicone granuloma have been reported in the literature,2,5,9,10 the dynamic MRI findings of
0009-9260/$ - see front matter ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2009.05.013
1176
breast cancer in injected breasts has, to our knowledge, been rarely described. The purpose of this retrospective study was to evaluate the diagnostic performance of dynamic MRI for breast cancer in breasts augmented with liquid paraffin or silicone injection.
Materials and methods This study was conducted with institutional review board approval and a waiver of patient informed consent because the study was retrospective. From January 1997 through December 2006, dynamic breast MRI was performed at our institution in 62 patients who had undergone breast augmentation with liquid paraffin or silicone injection. From them, patients were sought who had a pathological diagnosis or at least a 1-year MRI follow-up examination, and a total of 27 women (mean age 51.8 years; range 33e76 years) were found, constituting the study population. The injected foreign material was liquid paraffin in seven patients and liquid silicone in 20 patients. Eight patients had associated symptoms of palpable lumps. For patients who had undergone injection mammoplasty, mammography and/or sonography were used for the initial evaluation and if conventional images were BI-RADS category 0, dynamic breast MRI was recommended. Standard two-view mammography was obtained with dedicated equipment (Senograph DMR, GE, ilwaukee, WI, USA) and sonography was performed using a high-resolution unit with 10e12 MHz linear array transducer (HDI 5000 or 3000, Philips Advanced Technology Laboratories, Bothell, WA, USA). MRI was performed using a 1.5 T MRI system (Vision, Siemens, Erlangen, Germany) using a dedicated bilateral breast surface coil. Unenhanced images included a fat-suppressed, T2-weighted, turbo spin-echo sequence (repetition time/echo time (TR/TE) 4200 ms/ 90 ms; acquisition time 35 s; flip angle 180 ) and T1-weighted, conventional spin-echo pulse sequence (TR/TE 590 ms/12 ms; acquisition time 5 s; flip angle 90 ). Dynamic, T1-weighted, two-dimensional, fast low-angle shot sequences (TR/TE 32 ms/9 ms; flip angle 30 ) were acquired before and every 1 min for 5 min after an intravenous injection of 0.16 mmol gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) per kilogram of body weight. A section thickness of 3 mm, a full 256 256 imaging matrix, and a field of view adjusted to include both breasts (320 mm) was used. Subtraction images were obtained by subtracting the unenhanced images
J.H. Youk et al.
from all contrast-enhanced images. If contrast-enhanced lesions were identified on the subtraction images, region-of-interest-based timeesignal intensity curves were plotted to show the enhancement kinetics during the dynamic study. Two experienced breast radiologists who were blinded to the final diagnosis classified the morphological features and the enhancement kinetics of the enhancing lesions on MRI, in consensus. The morphological features were described using terminology according to the American College of Radiology BI-RADS11 and timeesignal intensity curve patterns were categorized into one of three types: a type 1 (persistent enhancement) pattern was assigned if the signal intensity increased steadily throughout the dynamic period; a type 2 (plateau) pattern was assigned if peak signal intensity was reached soon after the injection of the contrast medium and was followed by a plateau of signal intensity in the remaining dynamic series; a type 3 (washout) pattern was assigned if peak signal intensity was reached in the early phase and was immediately followed by a loss of signal intensity soon after the injection of the contrast medium.12 Readers were also asked to categorize the probability of malignancy for each enhancing lesion according to the BI-RADS final assessment. Those lesion characteristics at MRI were correlated with the final diagnosis based on the pathological result or at least 1-year MRI follow-up examination. As MRI-guided localization before surgical excision was not available in our institution, the gross finding at surgical investigation was correlated with the MRI findings to confirm whether the enhancing lesion at MRI was excised. Regarding malignant lesions, their pathological characteristics were reviewed. Also, the mammographic or sonographic findings of all patients were reviewed.
Results Of the 27 patients, seven patients underwent surgical excision and the remaining 20 patients underwent at least 1-year follow-up MRI (mean 2.8 years; range 1e7 years). At surgical excision, six invasive ductal carcinomas were diagnosed in six patients (age range 33e56 years; mean 49.2 years; Table 1) and one foreign body granuloma was diagnosed in one patient. Three of the six cancers had metastases in the ipsilateral axillary lymph nodes (Table 1). Of the other 20 patients, six patients had 10 enhancing lesions showing stability at follow-up MRI, regarded as benign and 14 patients had no enhancing lesion. Of eight patients with a palpable lump, five cancers and one benign
Diagnosis of breast cancer at dynamic MRI
1177
Table 1 Pathological characteristics of malignancy diagnosed at magnetic resonance imaging in patients with breast augmentation by paraffin or silicone injection No.
Size (mm)
Histological type
Histological grade
Vascular invasion
Nodal status
1 2 3 4 5 6
57 20 36 39 46 45
IDC IDC IDC IDC IDC IDC
II I I I II I
() () () () () ()
(þ) () (þ) () (þ) ()
IDC, invasive ductal carcinoma.
lesion were diagnosed after surgical excision, and the remaining two patients had no evidence of malignancy on the 2-year MRI follow-up examination. Mammograms and/or sonograms were available in 22 of 27 patients. On mammography, all patients had dense breasts with parenchymal distortion and densities with opaque foreign materials and granulomas randomly distributed within the breasts. Sonography was also limited owing to diffuse acoustic shadowing by overwhelming foreign material-induced fibrosis and granulomas. However, no evidence of malignancy was seen on either imaging method in those patients. A total of 17 enhancing lesions (mean maximal diameter measured at MRI 31.4 mm; range 5e57 mm) in 13 patients were found on MRI. Table 2 summarizes the morphological features at MRI. All six cancers (mean maximal diameter measured at MRI 40.5 mm; range 20e57 mm) were enhanced and showed suspicious MRI findings (i.e., irregular shape, irregular margins, ductal or segmental distribution, heterogeneous or internal rim Table 2 Morphological features of the enhancing lesions on magnetic resonance imaging in patients with breast augmentation by paraffin or silicone injection Morphologic features according to BI-RADS lexicon Mass Lesions
No. of lesions Shape
Margin
Internal enhancement Non-mass No. of lesions lesions Distribution
Internal enhancement
Malignant Benign
5 0 0 0 5 0 2 3 0 5 0 1 Segmental 1 Regional 0 Diffuse 0 Homogeneous 1 Round Oval Lobular Irregular Smooth Irregular Spiculate Homogeneous Heterogeneous Rim
6 4 1 1 0 6 0 0 5 0 1 5 0 3 2 5
enhancement); five appeared as enhancing masses (Fig. 1) and one appeared as a non-mass-like enhancement (Fig. 2). For 11 benign lesions, 10 enhancing lesions showed benign-favouring MRI findings (i.e., non-irregular shape, smooth margins, regional or diffuse distribution, homogeneous internal enhancement; Fig. 3), but one lesion showed lobular, smooth, rim enhancement. Regarding the enhancement kinetics, five of six cancers showed a type 2 pattern (Fig. 2) and the other showed a type 3 pattern (Fig. 1). All benign lesions showed a type 1 pattern (Fig. 3). Regarding the BI-RADS final assessment at MRI, 20 patients were negative (category 1 in 14 patients with no enhancement, category 2 in six patients) and the other seven were positive (category 4 in two patients, and category 5 in five patients). One of two lesions assigned to BI-RADS category 4 and all of five lesions assigned to BI-RADS category 5 were confirmed to be malignancies.
Discussion In patients who have undergone liquid paraffin or silicone injection into the breast, it is very difficult to diagnose breast cancer on conventional examinations, especially early-stage cancer. On physical examination, patients usually present with hard, lumpy breasts, which may mimic advanced breast cancer.5 Moreover, on mammography, foreign bodies are manifested as multiple radio-opaque masses that are often associated with architectural distortion or calcifications, which can obscure coexisting malignancy.5,13 Owing to strong acoustic shadowing, the ability of sonography to evaluate breast cancer is also compromised.2 However, MRI can provide information about the anatomical composition and internal architecture without obscuration by foreign-body granuloma and the kinetic analysis of enhancement pattern after the administration of contrast material can help the characterization of visible abnormalities in the augmented breast.7 In this study, dynamic MRI was able to diagnose breast cancers successfully in patients whose breast had been augmented by foreign materials. All cancers showed typical malignant morphological and kinetic features. Furthermore, they could be differentiated from benign lesions at MRI. In the literature, the morphological features suspicious of malignancy at MRI in foreign material-injected breasts are a spiculate, enhancing mass; an illdefined, lobulate mass with heterogeneous enhancement; or a huge soft-tissue mass with heterogeneous enhancement that was mixed with a haematoma.4,8
1178
J.H. Youk et al.
Figure 1 A 56-year-old woman with breast augmentation by injection of liquid paraffin and a palpable mass in the lower outer quadrant of the left breast. (a) An axial subtracted image of two-dimensional, fast low-angle shot MRI shows a 36-mm, spiculate, heterogeneous mass (arrow) in the lower outer quadrant of the left breast. (b) The timeesignal intensity curve of this enhancing mass shows washout enhancement kinetics.
Regarding the timeesignal intensity curve pattern, a cancer in silicone-injected breast reported by Po et al.8 showed a type 3 pattern and three cancers in silicone-injected breasts reported by Cheung et al.2 showed a type 2 pattern, which is compatible with the present results. However, there can be overlap in the morphological or kinetic features of benign and malignant lesions. In patients with foreign material injection in the breast, the differentiation of a malignant lesion from a benign lesion is difficult because the breasts usually have foreign material-induced mastopathies. Thus, for successful diagnosis, all features of the lesion should be analysed together, i.e., kinetics and morphology of enhancement.14 The BI-RADS MRI lexicon can be useful for such a combined analysis because this system incorporates morphological and kinetic features of lesions identified at breast MRI, and indicates the subsequent management. As expected, the BI-RADS classification performed well in the
present study. All malignancies were assigned to BI-RADS category 4 or 5, so that immediate pathological diagnosis could be made. In lesions assigned to BI-RADS category 1 or 2, no malignancy was found. Regarding the stage of breast cancer at detection, MRI may be expected to contribute to the earlier detection of cancer. Unfortunately, however, the present results were unable to support this because all the malignancies were of T2 or T3 stage (range of maximal diameter measured at MRI 20e57 mm) and three had positive nodes in the ipsilateral axilla. This disappointing result may be because the MRI examination was employed for assessing symptomatic disease. In injected breasts, breast lumps are usual and can be ignored easily by patients or clinicians, so that cancers would be detected at an advanced stage. The observations of the present study were limited by the small sample size in this series and by the lack of pathological studies or long-term
Diagnosis of breast cancer at dynamic MRI
1179
Figure 2 A 33-year-old woman with breast augmentation by injection of liquid silicone. (a) Breast sonography shows diffuse acoustic shadowing, which limits the evaluation of the parenchyma. (b) An axial subtracted image of two-dimensional, fast low-angle shot MRI shows a 46-mm segmental enhancement (arrow) in the outer area of the left breast. (c) The timeesignal intensity curve of this enhancing lesion shows plateau enhancement kinetics.
follow-up for some enhancing lesions at MRI. Additional long-term studies with more patients would be contributory. For the evaluation of breast abnormalities in patients whose breasts have been augmented
using paraffin or silicone injection, conventional imaging, such as mammography or sonography, is very limited. However, using dynamic breast MRI, malignant breast lesions were diagnosed successfully. Dynamic breast MRI with incorporated
1180
J.H. Youk et al.
Figure 3 A 45-year-old woman with breast augmentation by injection of liquid silicone. (a) Mammography shows multiple, hyperdense lesions of opaque foreign materials and granulomas diffusely distributed within the breasts. (b) An axial subtracted image of two-dimensional, fast low-angle shot MRI shows two oval, smooth, homogeneous masses (arrows) in the left breast. (c) The timeesignal intensity curve of the lateral lesion shows persistent enhancement kinetics. (d) An axial subtracted image of two-dimensional, fast low-angle shot MRI performed 2 years later shows no enhancing lesion.
analysis of morphological and kinetic features can be useful for the diagnosis of malignancy.
References 1. Middleton MS, McNamara Jr MP. A history and overview of breast augmentation and implant imaging. In: Middleton MS, editor. Breast implant imaging. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 1e20. 2. Cheung YC, Su MY, Ng SH, et al. Lumpy silicone-injected breasts: enhanced MRI and microscopic correlation. Clin Imaging 2002;26:397e404. 3. Erguvan-Dogan B, Yang WT. Direct injection of paraffin into the breast: mammographic, sonographic, and MRI features of early complications. AJR Am J Roentgenol 2006;186:888e94. 4. Peng HL, Wu CC, Choi WM, et al. Breast cancer detection using magnetic resonance imaging in breasts injected with liquid silicone. Plast Reconstr Surg 1999;104:2116e20. 5. Khong PL, Ho LW, Chan JH, et al. MR imaging of breast paraffinomas. AJR Am J Roentgenol 1999;173:929e32. 6. Sinclair DS, Olsen J, Spigos DG. Breast case of the day. AJR Am J Roentgenol 2000;175:859e61.
7. Cheung YC, Lee KF, Ng SH, et al. Sonographic features with histologic correlation in two cases of palpable breast cancer after breast augmentation by liquid silicone injection. J Clin Ultrasound 2002;30:548e51. 8. Po J, Margolis DJ, Cunningham CH, et al. Water-selective spectral-spatial contrast-enhanced breast MRI for cancer detection in patients with extracapsular and injected free silicone. Magn Reson Imaging 2006;24:1363e7. 9. Helbich TH, Wunderbaldinger P, Plenk H, et al. The value of MRI in silicone granuloma of the breast. Eur J Radiol 1997; 24:155e8. 10. Wang J, Shih TT, Li YW, et al. Magnetic resonance imaging characteristics of paraffinomas and siliconomas after mammoplasty. J Formos Med Assoc 2002;101:117e23. 11. American College of Radiology. Breast imaging reporting and data system, breast imaging atlas. 4th ed. Reston: American College of Radiology; 2003. 12. 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 1999;211:101e10. 13. Kim KT, Kim JH, Han BK, et al. MR findings of siliconoma in interstitial silicone injection mammoplasty patients. J Korean Radiol Soc 2002;46:73e9. 14. Morris EA. Breast MR imaging lexicon updated. Magn Reson Imaging Clin N Am 2006;14:293e303.
ORIGINAL PAPER
Diagnosis of breast cancer at dynamic MRI in patients with breast augmentation by paraffin or silicone injection J.H. Youk, E.J. Son*, E.-K. Kim, J.-A. Kim, M.J. Kim, J.Y. Kwak, S.M. Lee Department of Radiology, Research Institute of Radiological Science, Yonsei University College of Medicine, Seoul, Republic of Korea Received 10 February 2009; received in revised form 30 April 2009; accepted 11 May 2009
AIM: To determine the diagnostic performance of dynamic magnetic resonance imaging (MRI) for breast cancer in breasts augmented with liquid paraffin or silicone injection. MATERIALS AND METHODS: Among 62 patients with breast augmentation by liquid paraffin or silicone injection who had undergone dynamic breast MRI at our institution, 27 women, who had pathological diagnosis or at least 1-year MRI follow-up, were included in this retrospective study and their MRI images were reviewed. For enhancing lesions on MRI, the morphological features, enhancement kinetics, and BI-RADS assessment category were analysed. The lesion characteristics at MRI were correlated with the final diagnosis based on the histopathological result or at least 1-year MRI follow-up. RESULTS: Of the 27 patients, 17 enhancing lesions in 13 patients were found on MRI. All six lesions that were confirmed as malignancy showed suspicious morphological findings and type 2 or 3 enhancement kinetics, assigned to BIRADS category 4 or 5. Of the remaining 11 benign lesions, 10 showed benign-favouring morphological findings, and all showed type 1 enhancement kinetics, assigned to BI-RADS category 2 or 4. CONCLUSION: In patients with breasts injected with foreign material, MRI was used to successfully diagnose malignant breast lesions and could be the diagnostic method of choice. Analysis of the morphological and kinetic features at MRI in conjunction with clinical findings is essential. ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction The direct injection of foreign material, such as liquid paraffin and liquid silicone, into the breast was first used for breast augmentation in the early 1900s for paraffin and the 1960s for silicone.1 However, it has been largely abandoned owing to serious complications associated with the procedure, including painful nodularity, skin sloughing, migration of materials, granulomatous reactions, embolism, and even death.2 Nevertheless, this procedure has been performed illicitly by non-physicians and has appealed * Guarantor and correspondent: E. J. Son, Department of Radiology, Gangnam Severance Hospital, 146-92, Dogok-Dong, Kangnam-Ku, Seoul 135-720, Republic of Korea. Tel.: þ82 2 2019 3510; fax: þ82 2 3462 5472. E-mail address: [email protected] (E.J. Son).
to those who were not knowledgeable about its risks and complications because it is cheap, quick, and simple to perform.3,4 In the breast, the injected foreign materials cause a foreign-body granulomatous reaction and fibrosis with hard, nodular breast masses and architectural distortion mimicking neoplasm.5,6 Therefore, the role of conventional methods, such as physical, mammographic, and sonographic examination, is limited for the diagnosis of breast cancer in patients with breasts augmented by foreign materials, which causes a delay in the diagnosis.2 Dynamic contrast-enhanced magnetic resonance imaging (MRI) is thought to be an appropriate method for evaluation of lesions in such injected breasts, by virtue of its high temporal and spatial resolution.2,4,5,7,8 Although MRI features of breast paraffinoma or silicone granuloma have been reported in the literature,2,5,9,10 the dynamic MRI findings of
0009-9260/$ - see front matter ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2009.05.013
1176
breast cancer in injected breasts has, to our knowledge, been rarely described. The purpose of this retrospective study was to evaluate the diagnostic performance of dynamic MRI for breast cancer in breasts augmented with liquid paraffin or silicone injection.
Materials and methods This study was conducted with institutional review board approval and a waiver of patient informed consent because the study was retrospective. From January 1997 through December 2006, dynamic breast MRI was performed at our institution in 62 patients who had undergone breast augmentation with liquid paraffin or silicone injection. From them, patients were sought who had a pathological diagnosis or at least a 1-year MRI follow-up examination, and a total of 27 women (mean age 51.8 years; range 33e76 years) were found, constituting the study population. The injected foreign material was liquid paraffin in seven patients and liquid silicone in 20 patients. Eight patients had associated symptoms of palpable lumps. For patients who had undergone injection mammoplasty, mammography and/or sonography were used for the initial evaluation and if conventional images were BI-RADS category 0, dynamic breast MRI was recommended. Standard two-view mammography was obtained with dedicated equipment (Senograph DMR, GE, ilwaukee, WI, USA) and sonography was performed using a high-resolution unit with 10e12 MHz linear array transducer (HDI 5000 or 3000, Philips Advanced Technology Laboratories, Bothell, WA, USA). MRI was performed using a 1.5 T MRI system (Vision, Siemens, Erlangen, Germany) using a dedicated bilateral breast surface coil. Unenhanced images included a fat-suppressed, T2-weighted, turbo spin-echo sequence (repetition time/echo time (TR/TE) 4200 ms/ 90 ms; acquisition time 35 s; flip angle 180 ) and T1-weighted, conventional spin-echo pulse sequence (TR/TE 590 ms/12 ms; acquisition time 5 s; flip angle 90 ). Dynamic, T1-weighted, two-dimensional, fast low-angle shot sequences (TR/TE 32 ms/9 ms; flip angle 30 ) were acquired before and every 1 min for 5 min after an intravenous injection of 0.16 mmol gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) per kilogram of body weight. A section thickness of 3 mm, a full 256 256 imaging matrix, and a field of view adjusted to include both breasts (320 mm) was used. Subtraction images were obtained by subtracting the unenhanced images
J.H. Youk et al.
from all contrast-enhanced images. If contrast-enhanced lesions were identified on the subtraction images, region-of-interest-based timeesignal intensity curves were plotted to show the enhancement kinetics during the dynamic study. Two experienced breast radiologists who were blinded to the final diagnosis classified the morphological features and the enhancement kinetics of the enhancing lesions on MRI, in consensus. The morphological features were described using terminology according to the American College of Radiology BI-RADS11 and timeesignal intensity curve patterns were categorized into one of three types: a type 1 (persistent enhancement) pattern was assigned if the signal intensity increased steadily throughout the dynamic period; a type 2 (plateau) pattern was assigned if peak signal intensity was reached soon after the injection of the contrast medium and was followed by a plateau of signal intensity in the remaining dynamic series; a type 3 (washout) pattern was assigned if peak signal intensity was reached in the early phase and was immediately followed by a loss of signal intensity soon after the injection of the contrast medium.12 Readers were also asked to categorize the probability of malignancy for each enhancing lesion according to the BI-RADS final assessment. Those lesion characteristics at MRI were correlated with the final diagnosis based on the pathological result or at least 1-year MRI follow-up examination. As MRI-guided localization before surgical excision was not available in our institution, the gross finding at surgical investigation was correlated with the MRI findings to confirm whether the enhancing lesion at MRI was excised. Regarding malignant lesions, their pathological characteristics were reviewed. Also, the mammographic or sonographic findings of all patients were reviewed.
Results Of the 27 patients, seven patients underwent surgical excision and the remaining 20 patients underwent at least 1-year follow-up MRI (mean 2.8 years; range 1e7 years). At surgical excision, six invasive ductal carcinomas were diagnosed in six patients (age range 33e56 years; mean 49.2 years; Table 1) and one foreign body granuloma was diagnosed in one patient. Three of the six cancers had metastases in the ipsilateral axillary lymph nodes (Table 1). Of the other 20 patients, six patients had 10 enhancing lesions showing stability at follow-up MRI, regarded as benign and 14 patients had no enhancing lesion. Of eight patients with a palpable lump, five cancers and one benign
Diagnosis of breast cancer at dynamic MRI
1177
Table 1 Pathological characteristics of malignancy diagnosed at magnetic resonance imaging in patients with breast augmentation by paraffin or silicone injection No.
Size (mm)
Histological type
Histological grade
Vascular invasion
Nodal status
1 2 3 4 5 6
57 20 36 39 46 45
IDC IDC IDC IDC IDC IDC
II I I I II I
() () () () () ()
(þ) () (þ) () (þ) ()
IDC, invasive ductal carcinoma.
lesion were diagnosed after surgical excision, and the remaining two patients had no evidence of malignancy on the 2-year MRI follow-up examination. Mammograms and/or sonograms were available in 22 of 27 patients. On mammography, all patients had dense breasts with parenchymal distortion and densities with opaque foreign materials and granulomas randomly distributed within the breasts. Sonography was also limited owing to diffuse acoustic shadowing by overwhelming foreign material-induced fibrosis and granulomas. However, no evidence of malignancy was seen on either imaging method in those patients. A total of 17 enhancing lesions (mean maximal diameter measured at MRI 31.4 mm; range 5e57 mm) in 13 patients were found on MRI. Table 2 summarizes the morphological features at MRI. All six cancers (mean maximal diameter measured at MRI 40.5 mm; range 20e57 mm) were enhanced and showed suspicious MRI findings (i.e., irregular shape, irregular margins, ductal or segmental distribution, heterogeneous or internal rim Table 2 Morphological features of the enhancing lesions on magnetic resonance imaging in patients with breast augmentation by paraffin or silicone injection Morphologic features according to BI-RADS lexicon Mass Lesions
No. of lesions Shape
Margin
Internal enhancement Non-mass No. of lesions lesions Distribution
Internal enhancement
Malignant Benign
5 0 0 0 5 0 2 3 0 5 0 1 Segmental 1 Regional 0 Diffuse 0 Homogeneous 1 Round Oval Lobular Irregular Smooth Irregular Spiculate Homogeneous Heterogeneous Rim
6 4 1 1 0 6 0 0 5 0 1 5 0 3 2 5
enhancement); five appeared as enhancing masses (Fig. 1) and one appeared as a non-mass-like enhancement (Fig. 2). For 11 benign lesions, 10 enhancing lesions showed benign-favouring MRI findings (i.e., non-irregular shape, smooth margins, regional or diffuse distribution, homogeneous internal enhancement; Fig. 3), but one lesion showed lobular, smooth, rim enhancement. Regarding the enhancement kinetics, five of six cancers showed a type 2 pattern (Fig. 2) and the other showed a type 3 pattern (Fig. 1). All benign lesions showed a type 1 pattern (Fig. 3). Regarding the BI-RADS final assessment at MRI, 20 patients were negative (category 1 in 14 patients with no enhancement, category 2 in six patients) and the other seven were positive (category 4 in two patients, and category 5 in five patients). One of two lesions assigned to BI-RADS category 4 and all of five lesions assigned to BI-RADS category 5 were confirmed to be malignancies.
Discussion In patients who have undergone liquid paraffin or silicone injection into the breast, it is very difficult to diagnose breast cancer on conventional examinations, especially early-stage cancer. On physical examination, patients usually present with hard, lumpy breasts, which may mimic advanced breast cancer.5 Moreover, on mammography, foreign bodies are manifested as multiple radio-opaque masses that are often associated with architectural distortion or calcifications, which can obscure coexisting malignancy.5,13 Owing to strong acoustic shadowing, the ability of sonography to evaluate breast cancer is also compromised.2 However, MRI can provide information about the anatomical composition and internal architecture without obscuration by foreign-body granuloma and the kinetic analysis of enhancement pattern after the administration of contrast material can help the characterization of visible abnormalities in the augmented breast.7 In this study, dynamic MRI was able to diagnose breast cancers successfully in patients whose breast had been augmented by foreign materials. All cancers showed typical malignant morphological and kinetic features. Furthermore, they could be differentiated from benign lesions at MRI. In the literature, the morphological features suspicious of malignancy at MRI in foreign material-injected breasts are a spiculate, enhancing mass; an illdefined, lobulate mass with heterogeneous enhancement; or a huge soft-tissue mass with heterogeneous enhancement that was mixed with a haematoma.4,8
1178
J.H. Youk et al.
Figure 1 A 56-year-old woman with breast augmentation by injection of liquid paraffin and a palpable mass in the lower outer quadrant of the left breast. (a) An axial subtracted image of two-dimensional, fast low-angle shot MRI shows a 36-mm, spiculate, heterogeneous mass (arrow) in the lower outer quadrant of the left breast. (b) The timeesignal intensity curve of this enhancing mass shows washout enhancement kinetics.
Regarding the timeesignal intensity curve pattern, a cancer in silicone-injected breast reported by Po et al.8 showed a type 3 pattern and three cancers in silicone-injected breasts reported by Cheung et al.2 showed a type 2 pattern, which is compatible with the present results. However, there can be overlap in the morphological or kinetic features of benign and malignant lesions. In patients with foreign material injection in the breast, the differentiation of a malignant lesion from a benign lesion is difficult because the breasts usually have foreign material-induced mastopathies. Thus, for successful diagnosis, all features of the lesion should be analysed together, i.e., kinetics and morphology of enhancement.14 The BI-RADS MRI lexicon can be useful for such a combined analysis because this system incorporates morphological and kinetic features of lesions identified at breast MRI, and indicates the subsequent management. As expected, the BI-RADS classification performed well in the
present study. All malignancies were assigned to BI-RADS category 4 or 5, so that immediate pathological diagnosis could be made. In lesions assigned to BI-RADS category 1 or 2, no malignancy was found. Regarding the stage of breast cancer at detection, MRI may be expected to contribute to the earlier detection of cancer. Unfortunately, however, the present results were unable to support this because all the malignancies were of T2 or T3 stage (range of maximal diameter measured at MRI 20e57 mm) and three had positive nodes in the ipsilateral axilla. This disappointing result may be because the MRI examination was employed for assessing symptomatic disease. In injected breasts, breast lumps are usual and can be ignored easily by patients or clinicians, so that cancers would be detected at an advanced stage. The observations of the present study were limited by the small sample size in this series and by the lack of pathological studies or long-term
Diagnosis of breast cancer at dynamic MRI
1179
Figure 2 A 33-year-old woman with breast augmentation by injection of liquid silicone. (a) Breast sonography shows diffuse acoustic shadowing, which limits the evaluation of the parenchyma. (b) An axial subtracted image of two-dimensional, fast low-angle shot MRI shows a 46-mm segmental enhancement (arrow) in the outer area of the left breast. (c) The timeesignal intensity curve of this enhancing lesion shows plateau enhancement kinetics.
follow-up for some enhancing lesions at MRI. Additional long-term studies with more patients would be contributory. For the evaluation of breast abnormalities in patients whose breasts have been augmented
using paraffin or silicone injection, conventional imaging, such as mammography or sonography, is very limited. However, using dynamic breast MRI, malignant breast lesions were diagnosed successfully. Dynamic breast MRI with incorporated
1180
J.H. Youk et al.
Figure 3 A 45-year-old woman with breast augmentation by injection of liquid silicone. (a) Mammography shows multiple, hyperdense lesions of opaque foreign materials and granulomas diffusely distributed within the breasts. (b) An axial subtracted image of two-dimensional, fast low-angle shot MRI shows two oval, smooth, homogeneous masses (arrows) in the left breast. (c) The timeesignal intensity curve of the lateral lesion shows persistent enhancement kinetics. (d) An axial subtracted image of two-dimensional, fast low-angle shot MRI performed 2 years later shows no enhancing lesion.
analysis of morphological and kinetic features can be useful for the diagnosis of malignancy.
References 1. Middleton MS, McNamara Jr MP. A history and overview of breast augmentation and implant imaging. In: Middleton MS, editor. Breast implant imaging. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 1e20. 2. Cheung YC, Su MY, Ng SH, et al. Lumpy silicone-injected breasts: enhanced MRI and microscopic correlation. Clin Imaging 2002;26:397e404. 3. Erguvan-Dogan B, Yang WT. Direct injection of paraffin into the breast: mammographic, sonographic, and MRI features of early complications. AJR Am J Roentgenol 2006;186:888e94. 4. Peng HL, Wu CC, Choi WM, et al. Breast cancer detection using magnetic resonance imaging in breasts injected with liquid silicone. Plast Reconstr Surg 1999;104:2116e20. 5. Khong PL, Ho LW, Chan JH, et al. MR imaging of breast paraffinomas. AJR Am J Roentgenol 1999;173:929e32. 6. Sinclair DS, Olsen J, Spigos DG. Breast case of the day. AJR Am J Roentgenol 2000;175:859e61.
7. Cheung YC, Lee KF, Ng SH, et al. Sonographic features with histologic correlation in two cases of palpable breast cancer after breast augmentation by liquid silicone injection. J Clin Ultrasound 2002;30:548e51. 8. Po J, Margolis DJ, Cunningham CH, et al. Water-selective spectral-spatial contrast-enhanced breast MRI for cancer detection in patients with extracapsular and injected free silicone. Magn Reson Imaging 2006;24:1363e7. 9. Helbich TH, Wunderbaldinger P, Plenk H, et al. The value of MRI in silicone granuloma of the breast. Eur J Radiol 1997; 24:155e8. 10. Wang J, Shih TT, Li YW, et al. Magnetic resonance imaging characteristics of paraffinomas and siliconomas after mammoplasty. J Formos Med Assoc 2002;101:117e23. 11. American College of Radiology. Breast imaging reporting and data system, breast imaging atlas. 4th ed. Reston: American College of Radiology; 2003. 12. 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 1999;211:101e10. 13. Kim KT, Kim JH, Han BK, et al. MR findings of siliconoma in interstitial silicone injection mammoplasty patients. J Korean Radiol Soc 2002;46:73e9. 14. Morris EA. Breast MR imaging lexicon updated. Magn Reson Imaging Clin N Am 2006;14:293e303.