- Email: [email protected]
Arterial spin labelling perfusion MRI of breast cancer using FAIR TrueFISP: Initial results
Clinical Radiology 68 (2013) e123ee127
Contents lists available at SciVerse ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Arterial spin labelling perfusion MRI of breast cancer using FAIR TrueFISP: Initial results S. Buchbender a, S. Obenauer a, S. Mohrmann b, P. Martirosian c, C. Buchbender a, F.R. Miese a, H.J. Wittsack a, M. Miekley a, G. Antoch a, R.S. Lanzman a, * a
Department of Diagnostic and Interventional Radiology, University of Dusseldorf, Medical Faculty, Dusseldorf, Germany Department of Gynecology and Obstetrics, University of Dusseldorf, Medical Faculty, Dusseldorf, Germany c € bingen, Tu € bingen, Section on Experimental Radiology, Department of Diagnostic and Interventional Radiology, University Hospital Tu Germany b
art icl e i nformat ion Article history: Received 7 June 2012 Received in revised form 8 October 2012 Accepted 11 October 2012
AIM: To assess the feasibility of an unenhanced, flow-sensitive, alternating inversion recovery-balanced steady-state free precession (FAIR TrueFISP) arterial spin labelling (ASL) magnetic resonance imaging (MRI) technique for quantification of breast cancer perfusion. MATERIALS AND METHODS: Eighteen untreated breast tumour patients (mean age 53 17 years, range 30e68 years) and four healthy controls (mean age 51 14 years, range 33e68 years) were enrolled in this study and were imaged using a clinical 1.5 T MRI machine. Perfusion measurements were performed using a coronal single-section ASL FAIR TrueFISP technique in addition to a routine breast MRI examination. T1 relaxation time of normal breast parenchyma was determined in four healthy volunteers using the variable flip angle approach. The definitive diagnosis was obtained at histology after biopsy or surgery and was available for all patients. RESULTS: ASL perfusion was successfully acquired in 13 of 18 tumour patients and in all healthy controls. The mean ASL perfusion of invasive ductal carcinoma tissue was significantly higher (88.2 39.5 ml/100 g/min) compared to ASL perfusion of normal breast parenchyma (24.9 12.7 ml/100 g/min; p < 0.05) and invasive lobular carcinoma (30.5 4.3 ml/100 g/min; p < 0.05). No significant difference was found between the mean ASL perfusion of normal breast parenchyma and invasive lobular carcinoma tissue (p ¼ 0.97). CONCLUSION: ASL MRI enables quantification of breast cancer perfusion without the use of contrast material. However, its impact on diagnosis and therapy management of breast tumours has to be evaluated in larger patient studies. Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction
* Guarantor and correspondent: R.S. Lanzman, University Dusseldorf, Medical Faculty, Department of Diagnostic and Interventional Radiology, Moorenstrasse 5, 40225 Dusseldorf, Germany. Tel.: þ49 211 811 7752; fax: þ49 211 811 6145. E-mail addresses: [email protected], [email protected] (R.S. Lanzman).
Magnetic resonance imaging (MRI) has emerged as a pivotal imaging tool for detection and characterization of breast tumours.1 MRI has the highest sensitivity of all imaging methods for the detection of breast cancer.2 Its specificity has been reported to be variable, particularly in early publications, but with increased experience of the reader a specificity of up to 99% can be achieved.2e5
0009-9260/$ e see front matter Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.crad.2012.10.011
e124
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
In order to improve lesion characterization, functional imaging techniques such as diffusion-weighted imaging (DWI) and T1-weighted dynamic contrast-enhanced MRI (DCE-MRI) are used. DWI assesses the diffusion of water molecules, which is related to the cellularity and microstructure of the tissue of interest.6 DCE-MRI is generally used to measure signal variations in breast lesions after injection of a paramagnetic contrast material. Descriptive analysis of the wash-in and wash-out curves of the contrast agent is an important criterion for the detection and differentiation of suspicious breast masses.7 Moreover, DCE-MRI allows for calculation of quantitative parameters.1 However, the quantification of tissue perfusion is based on different compartmental models in order to account for the contribution of both perfusion and vascular permeability to tissue enhancement. Perfusion is a marker of tumour vascularity, which is of particular interest in breast cancer. Previous studies indicate that tumour vascularity might correlate with prognosis.8 Novel therapy options are based on the inhibition of tumour angiogenesis, and the clinical value of anti-angiogenic therapies in combination with cytostatic drugs is currently under investigation. Therapy monitoring with control of tumour size may be particularly inappropriate for anti-angiogenic therapy characterization, as most clinical studies suggest that the primary effects of anti-angiogenic agents are rather cytostatic than cytotoxic. Therefore, imaging methods that can reflect the antiangiogenic activity of novel drugs are required.9 Arterial spin labelling (ASL) MRI enables the determination of tissue perfusion without the use of contrast material. ASL MRI relies on the labelling of arterial blood supplying the tissue of interest by alteration of its longitudinal magnetization. As a non-enhanced imaging technique, ASL MRI enables repeated measurement without the risk of potential toxicity. ASL MRI has been successfully applied to improve disease detection and characterization in other areas, including the brain, pancreas, and kidney.10e13 Noguchi et al.14 demonstrated the feasibility of ASL MRI for differentiation of brain tumours. Furthermore, recent studies have highlighted the potential for monitoring of anti-angiogenic therapy.9 The purpose of this pilot study was to investigate the feasibility of ASL MRI of the breast for quantification of perfusion of regular fibroglandular tissue and breast cancer using a flow-sensitive alternating inversion recovery balanced steady-state free precession (FAIR TrueFISP) ASL MRI technique.
MRI of the breast was performed using a 1.5 T system (Magnetom Avanto, Siemens AG, Healthcare Sector, Erlangen, Germany) with a dedicated bilateral four-channel breast coil. Patients were examined in the prone position. In addition to the routine clinical imaging protocol, the FAIR TrueFISP ASL technique introduced by Martirosian et al.15 was acquired in coronal orientation through the centre of the tumour based on unenhanced T2- and short tau inversion recovery (STIR)-sequences. With this technique, perfusion images were obtained by interleaved image acquisition after a non-selective and a sectionselective inversion pulse. Section-selective inversion was performed with a 10.24 ms adiabatic inversion radiofrequency (RF) FOCI (frequency offset corrected inversion) “T-shape” pulse with m ¼ 5, b ¼ 935, 1500 Hz bandwidth, and 20 mm section thickness. With an inversion time (TI) of 1200 ms the TrueFISP acquisition was performed with the following imaging parameters: 4.6 ms repetition time (TR), 2.3 ms echo time (TE), 70 flip angle, 128 128 matrix, 180 180 mm field of view (FOV), 651 Hz/pixel bandwidth, 8 mm section thickness. Fifty images with selective and 50 images with non-selective inversion were acquired. An additional proton density TrueFISP image without inversion was acquired for determination of the equilibrium magnetization, which is necessary for the calculation of perfusion. A centric reordered phase encoding was applied to achieve reliable quantitative perfusion maps. Total scan time for the FAIR TrueFISP sequence was 4.30 min. Measurements were performed in the four controls in the region of maximal breast glandular density in order to determine the T1 relaxation time of breast parenchyma. T1 relaxation times were determined using a T1-weighted three-dimensional (3D) FLASH (fast low-angle shot) sequence with varying flip angles of 12 and 20 . Initially, the anatomical detail and physiological accuracy of the perfusion maps were evaluated by visual inspection of the images. For further analysis, images were transferred to a stand-alone workstation. The bespoke software STROKETOOL V 2.4 (http://www.digitalimagesolutions.de) was used for perfusion calculation and T1 quantification. Normal breast parenchyma of controls displayed a mean relaxation time of 1043 ms, which is comparable to values reported in literature.16 The following equation (Eqn 1) was used for pixel-based calculation of perfusion maps: l DMðTIÞ TI (1) exp f ¼ 2TI M0 T1
Materials and methods This study was approved by the local ethics committee and informed consent was obtained from all participating individuals. Twenty-two individuals were enrolled in this study: 18 patients with suspected breast tumour (mean age 53 17 years, range 30e68 years) and four healthy controls (mean age 51 14 years, range 33e68 years). The definitive diagnosis was obtained at histology after biopsy or surgery and was available for all patients.
where M0 represents the tissue equilibrium magnetization per unit mass; f is the perfusion rate; and l the bloode tissue water partition coefficient, which is assumed to have a constant value of 80 ml/100 g. T1 was set to 1043 ms as described above. The breast tumour was defined on coloured perfusion maps (Fig 1) by a hand-drawn region of interest (ROI) by two radiologists (R.S.L., S.B.) in consensus. The total postprocessing time ranged between 6e9 min. Mean values
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
Figure 1 Box plot showing significantly higher ASL perfusion levels in invasive ductal carcinoma as compared to invasive lobular carcinoma and healthy breast parenchyma (p < 0.05).
and standard deviation as well as statistical analysis were tabulated and analysed using the Statistical Package for Social Sciences (SPSS for Windows Package 15.0, Chicago, IL, USA). Analysis of variance (ANOVA) with a post-hoc Scheffe test was performed for the calculation of group differences in perfusion levels.
Results Acquisition of ASL perfusion was successfully performed in 13 of 18 tumour patients and in all four controls. ASL perfusion imaging failed in five tumour patients due to motion artefacts (n ¼ 1), technical problems with fat saturation (n ¼ 2), or because the tumour was not included in the single-section measurement (n ¼ 2). The mean ASL perfusion of breast parenchyma of controls was 24.9 12.7 ml/100 g/min. Of the 13 tumour patients with successful ASL acquisitions, nine patients had undergone breast biopsy 13 14 days before MRI. According to histology after tumour resection, 10 patients had invasive ductal carcinoma [G3 (n ¼ 4), G2 (n ¼ 6); mean tumour size at surgery 2.7 1 cm in greatest diameter] and three had invasive lobular carcinoma (all G2; mean tumour size at surgery 3.7 2.2 cm). The mean ASL perfusion of invasive ductal carcinoma (n ¼ 10) was significantly higher (88.2 39.5 ml/100 g/ min) as compared to ASL perfusion of normal breast parenchyma (p < 0.05). There was no significant difference between the mean ASL perfusion invasive lobular carcinoma (n ¼ 3; 30.5 4.3 ml/100 g/min) and normal breast parenchyma (p ¼ 0.97). The mean ASL perfusion of invasive ductal carcinoma was significantly higher than the mean ASL perfusion of invasive lobular carcinoma (p < 0.05; Fig 2).
Discussion Technological advancements in cancer imaging have led to a shift from pure morphological tumour assessment
e125
towards more sophisticated approaches for functional assessments of tissue characteristics.17 In the present study a FAIR TrueFISP ASL technique for non-enhanced perfusion measurements in breast cancer was successfully applied. The preliminary results show that invasive ductal carcinoma can be differentiated from normal breast parenchyma using ASL MRI. Moreover, the present data indicate that a sub-specification of breast tumour types (ductal versus lobular) might be feasible. Only a few studies have attempted to quantify perfusion of breast cancer with contrast-enhanced MRI.18,19 Similar to the present findings, a significantly higher perfusion was reported in breast cancer as compared to fibroglandular tissue. For example, Makkat et al.18 have reported a mean blood flow of 84 46 ml/100 mg/min for mainly invasive ductal breast cancer using quantitative DCE-MRI.18 Furthermore, in the single-bolus T1-DCE study based on pharmacokinetic modelling of contrast agent concentration reported by Brix et al.,19 10 untreated carcinomas had a mean perfusion of 61 ml/100 mg/min. However, in a review article Vaupel and Hockel20 stated that tumour blood flow in breast tumours could vary considerably between 8e80 ml/100 ml/min. These values are in good accordance with the ASL perfusion values determined in the present study. With ASL, MRI quantification of tissue perfusion can be performed easily as signal changes are proportional to blood flow. This might be of particular interest for the detection of early changes of tumour perfusion under neoadjuvant chemotherapy, which has become the standard of care for locally advanced breast cancer to downstage the disease.21,22 A major advantage of the ASL imaging is that administration of contrast material is not required. Hence, patients with contraindications for gadolinium-based contrast material can be imaged safely, for example, patients with renal insufficiency. Furthermore, with ASL MRI, quantification of tissue perfusion can be performed easily as signal changes are proportional to blood flow. These properties might be particularly attractive to display the vascularity of different subtypes of breast cancers and to identify potential biological markers suitable for specific targeted therapies.23 Neoangiogenesis is the basis for cancer proliferation.21 Several studies have attempted to improve characterization of tumour perfusion, not only to increase specificity for tumour detection, but also to monitor treatment efficacy.21,24 In fact, neoadjuvant chemotherapy has become the standard of care for locally advanced breast cancer to downstage the disease.21,22 Response to chemotherapeutic agents is assessed by combining physical examination, ultrasound, mammography, and MRI.21 However, changes in tumour morphology, especially tumour size, occur relatively late in the course of a treatment. Other biomarkers that allow an early prediction of tumour response to therapy and detect potential non-responder patients in order to use alternative therapies, may be advantageous in preventing patients from unnecessary toxicity.21,22 ASL MRI seems to be particularly suitable for therapy monitoring, as contrast material is not required. Therefore, image
e126
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
Figure 2 Breast MRI image showing an invasive ductal breast cancer in the posterior third of the right breast at the 9-o’clock position (aed). Colour map of ASL perfusion measurement shows higher perfusion at the rim of the tumour (in ml/100 g/min) (a). A hyperintense heterogeneous mass is displayed using STIR imaging in the coronal plane displays (b), and native T1-weighted FLASH 3D in axial plane (c). The contrastenhanced subtraction image reveals the tumour with rim enhancement (d).
acquisition can be repeated without concerns regarding cumulative doses of contrast material. This point might have a positive influence on patients’ acceptance to undergo repeated measurements. In a previous feasibility study, Kawashima et al. used an EPI (echo-planar imaging) approach for ASL MRI of breast cancer.25 However, EPI sequences are susceptible to magnetic field inhomogeneities, especially at boundaries of air and soft tissue, leading to artefacts such as image distortion.24 Therefore, so far, mainly half-Fourier single-shot turbo spin echo (HASTE) or balanced steady-state free-precession (bSSFP)-based approaches have been used for extracerebral ASL imaging.10,12,13 Furthermore, diffusion-weighted imaging studies in human breast cancer have demonstrated that image artefacts in breast imaging are more pronounced with EPI-sequences compared with HASTE-sequences.26 Hence, FAIR TrueFISP ASL might be more advantageous than EPIbased in human breast cancer. Several limitations of this pilot study should be acknowledged. The applied bSSFP based ASL sequence is a single-section sequence, which has to be acquired before
application of contrast material. Optimally the ASLmeasurement has to be positioned in the centre of the tumour. For this purpose the tumour has to be localized on the basis of an anatomical sequence, in this study by use of a coronal T2- and STIR sequences. Therefore, depending on tumour size and breast glandular density, the tumour might be missed or partially covered as occurred in two patients in the present study. It is expected that further development of multisection ASL perfusion might further improve the application of ASL perfusion in clinical practice. T1 values of healthy fibroglandular tissue were used for perfusion quantification. T1 relaxation times of breast lesions have not been studied so far and may vary from T1 times of healthy fibroglandular tissue and might lead to erroneous perfusion quantification. Future studies should address this issue. Furthermore, the small number of patients included in the present study represents a limitation. In conclusion, the ASL perfusion technique enables quantification of breast cancer perfusion, yielding perfusion values similar to those reported in the literature. However,
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
its impact on diagnosis and therapy management of breast tumours has to be evaluated in larger patient studies.
References 1. Sinha S, Sinha U. Recent advances in breast MRI and MRS. NMR Biomed 2009;22:3e16. 2. Kuhl C, Weigel S, Schrading S, et al. Prospective multicenter cohort study to refine management recommendations for women at elevated familial risk of breast cancer: the EVA trial. J Clin Oncol 2010;28: 1450e7. € brunner SH, Bick U, Bradley WG, et al. International 3. Heywang-Ko investigation of breast MRI: results of a multicentre study (11 sites) concerning diagnostic parameters for contrast-enhanced MRI based on 519 histopathologically correlated lesions. Eur Radiol 2001;11:531e46. 4. Yen YF, Han KF, Daniel BL, et al. Dynamic breast MRI with spiral trajectories: 3D versus 2D. J Magn Reson Imaging 2000;11:351e9. 5. Huang W, Fisher PR, Dulaimy K, et al. Detection of breast malignancy: diagnostic MR protocol for improved specificity. Radiology 2004; 232:585e91. 6. Arlinghaus LR, Li X, Rahman AR, et al. On the relationship between the apparent diffusion coefficient and extravascular extracellular volume fraction in human breast cancer. Magn Reson Imaging 2011;29: 630e8. 7. Chou CP, Wu MT, Chang HT, et al. Monitoring breast cancer response to neoadjuvant systemic chemotherapy using parametric contrastenhanced MRI: a pilot study. Acad Radiol 2007;14:561e73. 8. Sauer G, Deissler H. Angiogenesis: prognostic and therapeutic implications in gynecologic and breast malignancies. Curr Opin Obstet Gynecol 2003;15:45e9. 9. de Bazelaire C, Alsop DC, George D, et al. Magnetic resonance imagingmeasured blood flow change after antiangiogenic therapy with PTK787/ ZK 222584 correlates with clinical outcome in metastatic renal cell carcinoma. Clin Cancer Res 2008;14:5548e54. 10. Schraml C, Boss A, Martirosian P, et al. FAIR true-FISP perfusion imaging of the thyroid gland. J Magn Reson Imaging 2007;26:66e71. 11. Boss A, Martirosian P, Klose U, et al. FAIR-TrueFISP imaging of cerebral perfusion in areas of high magnetic susceptibility differences at 1.5 and 3 Tesla. J Magn Reson Imaging 2007;25:924e31. 12. Schraml C, Schwenzer NF, Martirosian P, et al. Perfusion imaging of the pancreas using an arterial spin labeling technique. J Magn Reson Imaging 2008;28:1459e65.
e127
13. Lanzman R, Wittsack H, Martirosian P, et al. Quantification of renal allograft perfusion using arterial spin labeling MRI: initial results. Eur Radiol 2010;20:1485e91. 14. Noguchi T, Yoshiura T, Hiwatashi A, et al. Perfusion imaging of brain tumors using arterial spin-labeling: correlation with histopathologic vascular density. AJNR Am J Neuroradiol 2008;29:688e93. 15. Martirosian P, Klose U, Mader I, et al. FAIR true-FISP perfusion imaging of the kidneys. Magn Reson Med 2004;51:353e61. 16. Delille JP, Slanetz PJ, Yeh ED, et al. Physiologic changes in breast magnetic resonance imaging during the menstrual cycle: perfusion imaging, signal enhancement, and influence of the T1 relaxation time of breast tissue. Breast J 2005;11:236e41. 17. Gore JC, Manning HC, Quarles CC, et al. Magnetic resonance in the era of molecular imaging of cancer. Magn Reson Imaging 2011;29:587e600. 18. Makkat S, Luypaert R, Sourbron S, et al. Quantification of perfusion and permeability in breast tumors with a deconvolution-based analysis of second-bolus T1-DCE data. J Magn Reson Imaging 2007;25:1159e67. 19. Brix G, Kiessling F, Lucht R, et al. Microcirculation and microvasculature in breast tumors: pharmacokinetic analysis of dynamic MR image series. Magn Reson Med 2004;52:420e9. 20. Vaupel P, Hockel M. Blood supply, oxygenation status and metabolic micromilieu of breast cancers: characterization and therapeutic relevance. Int J Oncol 2000;17:869e79. 21. Iacconi C, Giannelli M, Marini C, et al. The role of mean diffusivity (MD) as a predictive index of the response to chemotherapy in locally advanced breast cancer: a preliminary study. Eur Radiol 2010;20:303e8. 22. DeMartini W, Lehman C, Partridge S. Breast MRI for cancer detection and characterization: a review of evidence-based clinical applications. Acad Radiol 2008;15:408e16. 23. Li SP, Padhani AR, Makris A. Dynamic contrast-enhanced magnetic resonance imaging and blood oxygenation level-dependent magnetic resonance imaging for the assessment of changes in tumour biology with treatment. J Natl Cancer Inst Monogr 2011;2011:103e7. 24. Ah-See ML, Makris A, Taylor NJ, et al. Early changes in functional dynamic magnetic resonance imaging predict for pathologic response to neoadjuvant chemotherapy in primary breast cancer. Clin Cancer Res 2008;14:6580e9. 25. Kawashima M, Katada Y, Shukuya T, et al. MR perfusion imaging using the arterial spin labeling technique for breast cancer. J Magn Reson Imaging 2012;35:436e40. 26. Baltzer PA, Renz DM, Herrmann KH, et al. Diffusion-weighted imaging (DWI) in MR mammography (MRM): clinical comparison of echo planar imaging (EPI) and half-Fourier single-shot turbo spin echo (HASTE) diffusion techniques. Eur Radiol 2009;19:1612e20.
Contents lists available at SciVerse ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Arterial spin labelling perfusion MRI of breast cancer using FAIR TrueFISP: Initial results S. Buchbender a, S. Obenauer a, S. Mohrmann b, P. Martirosian c, C. Buchbender a, F.R. Miese a, H.J. Wittsack a, M. Miekley a, G. Antoch a, R.S. Lanzman a, * a
Department of Diagnostic and Interventional Radiology, University of Dusseldorf, Medical Faculty, Dusseldorf, Germany Department of Gynecology and Obstetrics, University of Dusseldorf, Medical Faculty, Dusseldorf, Germany c € bingen, Tu € bingen, Section on Experimental Radiology, Department of Diagnostic and Interventional Radiology, University Hospital Tu Germany b
art icl e i nformat ion Article history: Received 7 June 2012 Received in revised form 8 October 2012 Accepted 11 October 2012
AIM: To assess the feasibility of an unenhanced, flow-sensitive, alternating inversion recovery-balanced steady-state free precession (FAIR TrueFISP) arterial spin labelling (ASL) magnetic resonance imaging (MRI) technique for quantification of breast cancer perfusion. MATERIALS AND METHODS: Eighteen untreated breast tumour patients (mean age 53 17 years, range 30e68 years) and four healthy controls (mean age 51 14 years, range 33e68 years) were enrolled in this study and were imaged using a clinical 1.5 T MRI machine. Perfusion measurements were performed using a coronal single-section ASL FAIR TrueFISP technique in addition to a routine breast MRI examination. T1 relaxation time of normal breast parenchyma was determined in four healthy volunteers using the variable flip angle approach. The definitive diagnosis was obtained at histology after biopsy or surgery and was available for all patients. RESULTS: ASL perfusion was successfully acquired in 13 of 18 tumour patients and in all healthy controls. The mean ASL perfusion of invasive ductal carcinoma tissue was significantly higher (88.2 39.5 ml/100 g/min) compared to ASL perfusion of normal breast parenchyma (24.9 12.7 ml/100 g/min; p < 0.05) and invasive lobular carcinoma (30.5 4.3 ml/100 g/min; p < 0.05). No significant difference was found between the mean ASL perfusion of normal breast parenchyma and invasive lobular carcinoma tissue (p ¼ 0.97). CONCLUSION: ASL MRI enables quantification of breast cancer perfusion without the use of contrast material. However, its impact on diagnosis and therapy management of breast tumours has to be evaluated in larger patient studies. Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction
* Guarantor and correspondent: R.S. Lanzman, University Dusseldorf, Medical Faculty, Department of Diagnostic and Interventional Radiology, Moorenstrasse 5, 40225 Dusseldorf, Germany. Tel.: þ49 211 811 7752; fax: þ49 211 811 6145. E-mail addresses: [email protected], [email protected] (R.S. Lanzman).
Magnetic resonance imaging (MRI) has emerged as a pivotal imaging tool for detection and characterization of breast tumours.1 MRI has the highest sensitivity of all imaging methods for the detection of breast cancer.2 Its specificity has been reported to be variable, particularly in early publications, but with increased experience of the reader a specificity of up to 99% can be achieved.2e5
0009-9260/$ e see front matter Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.crad.2012.10.011
e124
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
In order to improve lesion characterization, functional imaging techniques such as diffusion-weighted imaging (DWI) and T1-weighted dynamic contrast-enhanced MRI (DCE-MRI) are used. DWI assesses the diffusion of water molecules, which is related to the cellularity and microstructure of the tissue of interest.6 DCE-MRI is generally used to measure signal variations in breast lesions after injection of a paramagnetic contrast material. Descriptive analysis of the wash-in and wash-out curves of the contrast agent is an important criterion for the detection and differentiation of suspicious breast masses.7 Moreover, DCE-MRI allows for calculation of quantitative parameters.1 However, the quantification of tissue perfusion is based on different compartmental models in order to account for the contribution of both perfusion and vascular permeability to tissue enhancement. Perfusion is a marker of tumour vascularity, which is of particular interest in breast cancer. Previous studies indicate that tumour vascularity might correlate with prognosis.8 Novel therapy options are based on the inhibition of tumour angiogenesis, and the clinical value of anti-angiogenic therapies in combination with cytostatic drugs is currently under investigation. Therapy monitoring with control of tumour size may be particularly inappropriate for anti-angiogenic therapy characterization, as most clinical studies suggest that the primary effects of anti-angiogenic agents are rather cytostatic than cytotoxic. Therefore, imaging methods that can reflect the antiangiogenic activity of novel drugs are required.9 Arterial spin labelling (ASL) MRI enables the determination of tissue perfusion without the use of contrast material. ASL MRI relies on the labelling of arterial blood supplying the tissue of interest by alteration of its longitudinal magnetization. As a non-enhanced imaging technique, ASL MRI enables repeated measurement without the risk of potential toxicity. ASL MRI has been successfully applied to improve disease detection and characterization in other areas, including the brain, pancreas, and kidney.10e13 Noguchi et al.14 demonstrated the feasibility of ASL MRI for differentiation of brain tumours. Furthermore, recent studies have highlighted the potential for monitoring of anti-angiogenic therapy.9 The purpose of this pilot study was to investigate the feasibility of ASL MRI of the breast for quantification of perfusion of regular fibroglandular tissue and breast cancer using a flow-sensitive alternating inversion recovery balanced steady-state free precession (FAIR TrueFISP) ASL MRI technique.
MRI of the breast was performed using a 1.5 T system (Magnetom Avanto, Siemens AG, Healthcare Sector, Erlangen, Germany) with a dedicated bilateral four-channel breast coil. Patients were examined in the prone position. In addition to the routine clinical imaging protocol, the FAIR TrueFISP ASL technique introduced by Martirosian et al.15 was acquired in coronal orientation through the centre of the tumour based on unenhanced T2- and short tau inversion recovery (STIR)-sequences. With this technique, perfusion images were obtained by interleaved image acquisition after a non-selective and a sectionselective inversion pulse. Section-selective inversion was performed with a 10.24 ms adiabatic inversion radiofrequency (RF) FOCI (frequency offset corrected inversion) “T-shape” pulse with m ¼ 5, b ¼ 935, 1500 Hz bandwidth, and 20 mm section thickness. With an inversion time (TI) of 1200 ms the TrueFISP acquisition was performed with the following imaging parameters: 4.6 ms repetition time (TR), 2.3 ms echo time (TE), 70 flip angle, 128 128 matrix, 180 180 mm field of view (FOV), 651 Hz/pixel bandwidth, 8 mm section thickness. Fifty images with selective and 50 images with non-selective inversion were acquired. An additional proton density TrueFISP image without inversion was acquired for determination of the equilibrium magnetization, which is necessary for the calculation of perfusion. A centric reordered phase encoding was applied to achieve reliable quantitative perfusion maps. Total scan time for the FAIR TrueFISP sequence was 4.30 min. Measurements were performed in the four controls in the region of maximal breast glandular density in order to determine the T1 relaxation time of breast parenchyma. T1 relaxation times were determined using a T1-weighted three-dimensional (3D) FLASH (fast low-angle shot) sequence with varying flip angles of 12 and 20 . Initially, the anatomical detail and physiological accuracy of the perfusion maps were evaluated by visual inspection of the images. For further analysis, images were transferred to a stand-alone workstation. The bespoke software STROKETOOL V 2.4 (http://www.digitalimagesolutions.de) was used for perfusion calculation and T1 quantification. Normal breast parenchyma of controls displayed a mean relaxation time of 1043 ms, which is comparable to values reported in literature.16 The following equation (Eqn 1) was used for pixel-based calculation of perfusion maps: l DMðTIÞ TI (1) exp f ¼ 2TI M0 T1
Materials and methods This study was approved by the local ethics committee and informed consent was obtained from all participating individuals. Twenty-two individuals were enrolled in this study: 18 patients with suspected breast tumour (mean age 53 17 years, range 30e68 years) and four healthy controls (mean age 51 14 years, range 33e68 years). The definitive diagnosis was obtained at histology after biopsy or surgery and was available for all patients.
where M0 represents the tissue equilibrium magnetization per unit mass; f is the perfusion rate; and l the bloode tissue water partition coefficient, which is assumed to have a constant value of 80 ml/100 g. T1 was set to 1043 ms as described above. The breast tumour was defined on coloured perfusion maps (Fig 1) by a hand-drawn region of interest (ROI) by two radiologists (R.S.L., S.B.) in consensus. The total postprocessing time ranged between 6e9 min. Mean values
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
Figure 1 Box plot showing significantly higher ASL perfusion levels in invasive ductal carcinoma as compared to invasive lobular carcinoma and healthy breast parenchyma (p < 0.05).
and standard deviation as well as statistical analysis were tabulated and analysed using the Statistical Package for Social Sciences (SPSS for Windows Package 15.0, Chicago, IL, USA). Analysis of variance (ANOVA) with a post-hoc Scheffe test was performed for the calculation of group differences in perfusion levels.
Results Acquisition of ASL perfusion was successfully performed in 13 of 18 tumour patients and in all four controls. ASL perfusion imaging failed in five tumour patients due to motion artefacts (n ¼ 1), technical problems with fat saturation (n ¼ 2), or because the tumour was not included in the single-section measurement (n ¼ 2). The mean ASL perfusion of breast parenchyma of controls was 24.9 12.7 ml/100 g/min. Of the 13 tumour patients with successful ASL acquisitions, nine patients had undergone breast biopsy 13 14 days before MRI. According to histology after tumour resection, 10 patients had invasive ductal carcinoma [G3 (n ¼ 4), G2 (n ¼ 6); mean tumour size at surgery 2.7 1 cm in greatest diameter] and three had invasive lobular carcinoma (all G2; mean tumour size at surgery 3.7 2.2 cm). The mean ASL perfusion of invasive ductal carcinoma (n ¼ 10) was significantly higher (88.2 39.5 ml/100 g/ min) as compared to ASL perfusion of normal breast parenchyma (p < 0.05). There was no significant difference between the mean ASL perfusion invasive lobular carcinoma (n ¼ 3; 30.5 4.3 ml/100 g/min) and normal breast parenchyma (p ¼ 0.97). The mean ASL perfusion of invasive ductal carcinoma was significantly higher than the mean ASL perfusion of invasive lobular carcinoma (p < 0.05; Fig 2).
Discussion Technological advancements in cancer imaging have led to a shift from pure morphological tumour assessment
e125
towards more sophisticated approaches for functional assessments of tissue characteristics.17 In the present study a FAIR TrueFISP ASL technique for non-enhanced perfusion measurements in breast cancer was successfully applied. The preliminary results show that invasive ductal carcinoma can be differentiated from normal breast parenchyma using ASL MRI. Moreover, the present data indicate that a sub-specification of breast tumour types (ductal versus lobular) might be feasible. Only a few studies have attempted to quantify perfusion of breast cancer with contrast-enhanced MRI.18,19 Similar to the present findings, a significantly higher perfusion was reported in breast cancer as compared to fibroglandular tissue. For example, Makkat et al.18 have reported a mean blood flow of 84 46 ml/100 mg/min for mainly invasive ductal breast cancer using quantitative DCE-MRI.18 Furthermore, in the single-bolus T1-DCE study based on pharmacokinetic modelling of contrast agent concentration reported by Brix et al.,19 10 untreated carcinomas had a mean perfusion of 61 ml/100 mg/min. However, in a review article Vaupel and Hockel20 stated that tumour blood flow in breast tumours could vary considerably between 8e80 ml/100 ml/min. These values are in good accordance with the ASL perfusion values determined in the present study. With ASL, MRI quantification of tissue perfusion can be performed easily as signal changes are proportional to blood flow. This might be of particular interest for the detection of early changes of tumour perfusion under neoadjuvant chemotherapy, which has become the standard of care for locally advanced breast cancer to downstage the disease.21,22 A major advantage of the ASL imaging is that administration of contrast material is not required. Hence, patients with contraindications for gadolinium-based contrast material can be imaged safely, for example, patients with renal insufficiency. Furthermore, with ASL MRI, quantification of tissue perfusion can be performed easily as signal changes are proportional to blood flow. These properties might be particularly attractive to display the vascularity of different subtypes of breast cancers and to identify potential biological markers suitable for specific targeted therapies.23 Neoangiogenesis is the basis for cancer proliferation.21 Several studies have attempted to improve characterization of tumour perfusion, not only to increase specificity for tumour detection, but also to monitor treatment efficacy.21,24 In fact, neoadjuvant chemotherapy has become the standard of care for locally advanced breast cancer to downstage the disease.21,22 Response to chemotherapeutic agents is assessed by combining physical examination, ultrasound, mammography, and MRI.21 However, changes in tumour morphology, especially tumour size, occur relatively late in the course of a treatment. Other biomarkers that allow an early prediction of tumour response to therapy and detect potential non-responder patients in order to use alternative therapies, may be advantageous in preventing patients from unnecessary toxicity.21,22 ASL MRI seems to be particularly suitable for therapy monitoring, as contrast material is not required. Therefore, image
e126
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
Figure 2 Breast MRI image showing an invasive ductal breast cancer in the posterior third of the right breast at the 9-o’clock position (aed). Colour map of ASL perfusion measurement shows higher perfusion at the rim of the tumour (in ml/100 g/min) (a). A hyperintense heterogeneous mass is displayed using STIR imaging in the coronal plane displays (b), and native T1-weighted FLASH 3D in axial plane (c). The contrastenhanced subtraction image reveals the tumour with rim enhancement (d).
acquisition can be repeated without concerns regarding cumulative doses of contrast material. This point might have a positive influence on patients’ acceptance to undergo repeated measurements. In a previous feasibility study, Kawashima et al. used an EPI (echo-planar imaging) approach for ASL MRI of breast cancer.25 However, EPI sequences are susceptible to magnetic field inhomogeneities, especially at boundaries of air and soft tissue, leading to artefacts such as image distortion.24 Therefore, so far, mainly half-Fourier single-shot turbo spin echo (HASTE) or balanced steady-state free-precession (bSSFP)-based approaches have been used for extracerebral ASL imaging.10,12,13 Furthermore, diffusion-weighted imaging studies in human breast cancer have demonstrated that image artefacts in breast imaging are more pronounced with EPI-sequences compared with HASTE-sequences.26 Hence, FAIR TrueFISP ASL might be more advantageous than EPIbased in human breast cancer. Several limitations of this pilot study should be acknowledged. The applied bSSFP based ASL sequence is a single-section sequence, which has to be acquired before
application of contrast material. Optimally the ASLmeasurement has to be positioned in the centre of the tumour. For this purpose the tumour has to be localized on the basis of an anatomical sequence, in this study by use of a coronal T2- and STIR sequences. Therefore, depending on tumour size and breast glandular density, the tumour might be missed or partially covered as occurred in two patients in the present study. It is expected that further development of multisection ASL perfusion might further improve the application of ASL perfusion in clinical practice. T1 values of healthy fibroglandular tissue were used for perfusion quantification. T1 relaxation times of breast lesions have not been studied so far and may vary from T1 times of healthy fibroglandular tissue and might lead to erroneous perfusion quantification. Future studies should address this issue. Furthermore, the small number of patients included in the present study represents a limitation. In conclusion, the ASL perfusion technique enables quantification of breast cancer perfusion, yielding perfusion values similar to those reported in the literature. However,
S. Buchbender et al. / Clinical Radiology 68 (2013) e123ee127
its impact on diagnosis and therapy management of breast tumours has to be evaluated in larger patient studies.
References 1. Sinha S, Sinha U. Recent advances in breast MRI and MRS. NMR Biomed 2009;22:3e16. 2. Kuhl C, Weigel S, Schrading S, et al. Prospective multicenter cohort study to refine management recommendations for women at elevated familial risk of breast cancer: the EVA trial. J Clin Oncol 2010;28: 1450e7. € brunner SH, Bick U, Bradley WG, et al. International 3. Heywang-Ko investigation of breast MRI: results of a multicentre study (11 sites) concerning diagnostic parameters for contrast-enhanced MRI based on 519 histopathologically correlated lesions. Eur Radiol 2001;11:531e46. 4. Yen YF, Han KF, Daniel BL, et al. Dynamic breast MRI with spiral trajectories: 3D versus 2D. J Magn Reson Imaging 2000;11:351e9. 5. Huang W, Fisher PR, Dulaimy K, et al. Detection of breast malignancy: diagnostic MR protocol for improved specificity. Radiology 2004; 232:585e91. 6. Arlinghaus LR, Li X, Rahman AR, et al. On the relationship between the apparent diffusion coefficient and extravascular extracellular volume fraction in human breast cancer. Magn Reson Imaging 2011;29: 630e8. 7. Chou CP, Wu MT, Chang HT, et al. Monitoring breast cancer response to neoadjuvant systemic chemotherapy using parametric contrastenhanced MRI: a pilot study. Acad Radiol 2007;14:561e73. 8. Sauer G, Deissler H. Angiogenesis: prognostic and therapeutic implications in gynecologic and breast malignancies. Curr Opin Obstet Gynecol 2003;15:45e9. 9. de Bazelaire C, Alsop DC, George D, et al. Magnetic resonance imagingmeasured blood flow change after antiangiogenic therapy with PTK787/ ZK 222584 correlates with clinical outcome in metastatic renal cell carcinoma. Clin Cancer Res 2008;14:5548e54. 10. Schraml C, Boss A, Martirosian P, et al. FAIR true-FISP perfusion imaging of the thyroid gland. J Magn Reson Imaging 2007;26:66e71. 11. Boss A, Martirosian P, Klose U, et al. FAIR-TrueFISP imaging of cerebral perfusion in areas of high magnetic susceptibility differences at 1.5 and 3 Tesla. J Magn Reson Imaging 2007;25:924e31. 12. Schraml C, Schwenzer NF, Martirosian P, et al. Perfusion imaging of the pancreas using an arterial spin labeling technique. J Magn Reson Imaging 2008;28:1459e65.
e127
13. Lanzman R, Wittsack H, Martirosian P, et al. Quantification of renal allograft perfusion using arterial spin labeling MRI: initial results. Eur Radiol 2010;20:1485e91. 14. Noguchi T, Yoshiura T, Hiwatashi A, et al. Perfusion imaging of brain tumors using arterial spin-labeling: correlation with histopathologic vascular density. AJNR Am J Neuroradiol 2008;29:688e93. 15. Martirosian P, Klose U, Mader I, et al. FAIR true-FISP perfusion imaging of the kidneys. Magn Reson Med 2004;51:353e61. 16. Delille JP, Slanetz PJ, Yeh ED, et al. Physiologic changes in breast magnetic resonance imaging during the menstrual cycle: perfusion imaging, signal enhancement, and influence of the T1 relaxation time of breast tissue. Breast J 2005;11:236e41. 17. Gore JC, Manning HC, Quarles CC, et al. Magnetic resonance in the era of molecular imaging of cancer. Magn Reson Imaging 2011;29:587e600. 18. Makkat S, Luypaert R, Sourbron S, et al. Quantification of perfusion and permeability in breast tumors with a deconvolution-based analysis of second-bolus T1-DCE data. J Magn Reson Imaging 2007;25:1159e67. 19. Brix G, Kiessling F, Lucht R, et al. Microcirculation and microvasculature in breast tumors: pharmacokinetic analysis of dynamic MR image series. Magn Reson Med 2004;52:420e9. 20. Vaupel P, Hockel M. Blood supply, oxygenation status and metabolic micromilieu of breast cancers: characterization and therapeutic relevance. Int J Oncol 2000;17:869e79. 21. Iacconi C, Giannelli M, Marini C, et al. The role of mean diffusivity (MD) as a predictive index of the response to chemotherapy in locally advanced breast cancer: a preliminary study. Eur Radiol 2010;20:303e8. 22. DeMartini W, Lehman C, Partridge S. Breast MRI for cancer detection and characterization: a review of evidence-based clinical applications. Acad Radiol 2008;15:408e16. 23. Li SP, Padhani AR, Makris A. Dynamic contrast-enhanced magnetic resonance imaging and blood oxygenation level-dependent magnetic resonance imaging for the assessment of changes in tumour biology with treatment. J Natl Cancer Inst Monogr 2011;2011:103e7. 24. Ah-See ML, Makris A, Taylor NJ, et al. Early changes in functional dynamic magnetic resonance imaging predict for pathologic response to neoadjuvant chemotherapy in primary breast cancer. Clin Cancer Res 2008;14:6580e9. 25. Kawashima M, Katada Y, Shukuya T, et al. MR perfusion imaging using the arterial spin labeling technique for breast cancer. J Magn Reson Imaging 2012;35:436e40. 26. Baltzer PA, Renz DM, Herrmann KH, et al. Diffusion-weighted imaging (DWI) in MR mammography (MRM): clinical comparison of echo planar imaging (EPI) and half-Fourier single-shot turbo spin echo (HASTE) diffusion techniques. Eur Radiol 2009;19:1612e20.