- Email: [email protected]
Whole-body MR imaging of bone marrow
European Journal of Radiology 55 (2005) 33–40
Whole-body MR imaging of bone marrow G.P. Schmidt ∗ , S.O. Schoenberg, M.F. Reiser, A. Baur-Melnyk Institute of Clinical Radiology, University Hospitals Munich/Grosshadern, LMU, Marchioninistr. 15, M¨unchen 81377, Germany Received 17 January 2005; received in revised form 20 January 2005; accepted 26 January 2005
Abstract In clinical routine, multimodality algorithms, including X-ray, computed tomography, scintigraphy and MRI, are used in case of suspected bone marrow malignancy. Skeletal scintigraphy is widely used to asses metastatic disease to the bone, CT is the technique of choice to assess criteria of osseous destruction and bone stability. MRI is the only imaging technique that allows direct visualization of bone marrow and its components with high spatial resolution. The combination of unenhanced T1-weighted-spin echo- and turbo-STIR-sequences have shown to be most useful for the detection of bone marrow abnormalities and are able to discriminate benign from malignant bone marrow changes. Originally, whole-body MRI bone marrow screening was performed in sequential scanning techniques of five body levels with time consuming coil rearrangement and repositioning of the patient. The introduction of a rolling platform mounted on top of a conventional MRI examination table faciliated whole-body MR imaging and, with the use of fast gradient echo, T1-weighted and STIR-imaging techniques, for the first time allowed whole-body imaging within less than one hour. With the development of parallel imaging techniques (PAT) in combination with global matrix coil concepts, acquisition time could be reduced substantially without compromises in spatial resolution, enabling the implementation of more complex and flexible examination protocols. Whole-body MRI represents a new alternative to the stepwise multimodality concept for the detection of metastatic disease, multiple myeloma and lymphoma of the bone with high diagnostic accuracy. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: MRI; Whole body; Parallel imaging; Bone marrow; Metastases
1. Introduction The skeletal system is a frequent target of metastatic disease and constitutes the third most common location of metastatic manifestations after the liver and the lung. The primary tumors with the highest incidence of skeletal metastases are prostate cancer in men and breast cancer in women, followed by thyroid, renal and lung carcinomas. These tumors account for approximately 80% of metastatic bone disease [3]. Early detection of bone metastases has an important impact on patient management and may contribute to prevent complications like pathological fractures. Magnetic resonance imaging is the only imaging technique that allows direct visualization of the bone marrow and its components. It is the most sensitive technique for the detection of bone marrow pathologies [1,2]. Due to its high ∗
Corresponding author. Tel.: +49 89 7095 0; fax: +49 89 7095 8832. E-mail address: [email protected] (G.P. Schmidt). 0720-048X/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2005.01.019
soft tissue contrast and spatial resolution, MRI has become the method of choice for bone marrow imaging [35]. The predominantly hematogenous spread of bone metastases explains its frequent distribution in active hematopoetic bone marrow, especially in the axial skeleton. However, it has been reported that up to 40% of bone metastases occur in the appendicular skeleton stressing the need for whole-body anatomic coverage [4]. 2. Diagnostic imaging modalities for whole-body bone marrow screening In clinical practice most commonly multimodality algorithms are used in case of suspected metastatic bone disease, including conventional X-ray, bone marrow- or skeletal scintigraphy, PET, CT and MRI. These examinations are performed in order to evaluate the presence and type of bone lesions, to assess their extent and localization and eventually guide potential biopsy. All these modalities
34
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
have different performances in terms of sensitivity and specificity. 2.1. MRI versus radiography/CT In patients with a positive bone scan or with local symptoms and pain, radiographs and CT are usually performed for verification. Plain radiographs have a low sensitivity in the detection of metastases compared to MRI or other imaging modalities [22,37–39]. This is due to the fact that a change of 30–50% in mineral density is needed before a metastatic lesion becomes visible on plain film [14]. Compared to bone scintigraphy, it has been reported that radiography can be inapparent in nearly half of true-positive lesions revealed in scintigraphy [38]. Lecouvet et al. compared the performance of MRI and radiography to detect focal bone lesions in advanced Multiple myeloma and showed a clear superiority of MRI in given anatomic areas, like the spine (76% versus 42% detected lesions) and pelvis (75% versus 46% detected lesions) [22]. In the detection of bone destructions CT is far more sensitive than radiography: Krahe et al. compared conventional radiographs and CT examinations of 112 patients with primary and secondary bone tumors of the spine [40]. Of 268 involved vertebrae identified in CT, 88% were identified by radiographs when the vertebral body was infiltrated, but only 66% when other parts of the vertebrae were affected. Of the intraspinal and paravertebral tumor extensions only 23% and 33% were correctly diagnosed by plain radiography. Moreover, CT also allows to assess paraosseous tumor extension yields and it is the image modality of choice to evaluate the extent of osteolysis and to assess stability and fracture risk, respectively [24,36,39]. However, it is not yet clear if multislice-CT is equal to MRI for the assessment of bony metastases. 2.2. MRI versus scintigraphy In asymptomatic patients with suspected skeletal metastases bone scintigraphy is well established. This widely available technique is based on “indirectly” detecting bone metastases by registration of tracer uptake in areas of osteoblastic or osteoclastic activity and has long been regarded as a standard of reference [5]. In various studies, the efficiency of MRI has been compared to bone scintigraphy or bone marrow scintigraphy for the detection of bony metastases [1,2,6,12,34]. In a most recent study Ghanem et al. examined 129 patients with malignant tumors and suspected bone metastases using both whole-body MRI (coronary turbo-STIR-imaging) and bone scintigraphy [34]. Whole-body MRI and skeletal scintigraphy demonstrated skeletal metastases in 49/129 patients with a concordance of 81%. In this study, whole-body MRI revealed skeletal metastases in 15 cases which had not been found in skeletal scintigraphy. Moreover, MRI showed a larger extent of bone infiltration. In 60% of the cases, whole-body MRI evidenced tumor-associated findings, such as lymph node metastases in 42% of cases or metastases in
parenchymal organs in 23%. The same study group compared the performance of bone marrow scintigraphy with wholebody MRI in 20 patients with skeletal metastases: in 60% of patients MRI revealed more metastases and local therapy (e.g. radiotherapy or surgery) was applied on 78% of these patients [1]. The superiority of MRI compared to skeletal and bone marrow scintigraphy in depicting bone metastases has also been reported by other authors [2,6,12]. The high performance of MRI is mainly due to the fact that MRI can visually depict metastatic bone marrow infiltration at an early stage, before osseous changes occur due to osteoblastic or osteoclastic activity. For an MRI bone screening, the combination of non-contrasted T1-weighted-Spinechoand Turbo-STIR-sequences proved to be most sensitive and it also allows to reliably discriminate benign from malignant marrow disorders [6,19,20]. On T1-weighted sequences tumor spread is identified by replacement of normal marrow, resulting in an isointense or hypointense signal compared to muscle tissue. On fat suppressed sequences, like STIR (short tau inversion time inversion recovery), neoplastic lesions are readily detected by virtue of the hyperintense signal due to an increased content of water within tumor cells, in contrast to the surrounding normal marrow. In osteoblastic metastases areas of low signal intensity on T1-weighted turbo-SE images correspond to areas of low signal intensity on T2weighted turbo-SE images. On STIR-sequences osteoblastic metastases show a spectrum from no signal change in very dense sclerotic metastases to hyperintense signal in cases where more cellular components do exist [35]. Especially bone marrow infiltration by tumor cells which is not associated with bone destruction or formation of new bone, can be detected with MRI. Moreover, MRI also enables to differentiate reactive sclerosis, e.g. following fractures, from neoplastic osteosclerosis. The unique soft-tissue contrast of MRI enables for precise assessment of tumor infiltration within the bone marrow and adjacent paraosseous structures, such as the spinal canal. Finally, MRI can help to define the localization of biopsy, if needed. MRI also is particularly helpful to monitor therapy response, e.g. after radiation therapy [28].
3. Methods for whole-body MR Imaging The main challenge for whole-body MRI has been to cope with substantial differences in coil requirements, slice positioning and orientation into one comprehensive scan. Repositioning of the patient as well as the coils has been time consuming. Therefore, attempts for whole-body MRI were always associated with compromises in spatial resolution and image quality. 3.1. Sequential scanning In order to cover the whole body with a scanner of conventional design (Magnetom Symphony, Siemens Medical Solutions) five levels had to be scanned separately. First, the
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
patient was examined head first with coronal scans of the skull/neck and the thorax using a combination of head coil, neck array coil and 1–2 body array coils. Sagittal images of the spine were acquired using the spine array coil. T1weighted-spinecho- (TR 92,2, TE 4,1) and STIR-sequences (TR 3912, TE 60, TI 150) with a maximum field of view of 450 mm were obtained. After repositioning the patient feet first, the pelvis, femoral bones and lower legs were imaged in coronal planes, using body array and spine array coils and the same sequence types. Due to the need to reposition the patient, the total room time for this protocol was approximately 60 min. Steinborn et al. examined 18 patients with suspected skeletal metastases and correlated results with bone scintigraphy [6]. Out of 105 proven metastatic lesions, 96 (91.4%) were detected with whole-body MRI, whereas 89 lesions (84.8%) were detected with bone scintigraphy. In addition, MRI allowed to detect liver and lung metastases which of course, were not depicted in the bone scans. Approximately 1/4 of confirmed malignant lesions (23.8%) would have been missed in a “classic” MRI screening protocol covering only the spine, pelvis and proximal femurs. Nine skeletal metastases detected in the bone scan but missed in MRI were located either in the ribcage or in the skull, so that it can be conchided that these anatomical areas are not well represented with the described whole MRI technique. 3.2. Rolling table platform The introduction of dedicated rolling platforms with a fixed surface coil was a first step towards performing wholebody MRI within a single examination and towards a significant reduction of examination times and an increase in patient comfort [7]. The setup consists of a rolling table (length 270 cm, width 33–50 cm) mounted on top of the scanner table allowing free movement in the z-direction. For signal acquisition, the spine array and body phased array, which remains fixed to the scanner in the centre of the gantry, are employed. With the use of the rolling platform, the patient can be ex-
35
amined from head to toe sliding in between this so called “coil sandwich” (Fig. 1). Barkhausen et al. presented this setup for the first time as a quick whole-body MRI screening method for metastatic disease. The protocol was based on the acquisition of axial real-time trueFISP-gradient echo sequences (TR 2,2, TE 1,1; flip angle 60%). Using echo sharing to improve temporal resolution, the total scanning time was brought down to 30 s [7]. All 11 bone metastases detected by CT as a reference method were identified with MRI. Thus, four lesions depicted in bone scintigraphy, mainly located in the ribcage, were missed with MRI. A disadvantage of rapid acquisition data sets, which are often derived from gradient echo- or echoplanar imaging techniques, is a high sensitivity to image distortion caused by magnetic susceptibility effects, especially at air/soft-tissue interfaces as well as poor soft tissue contrast. Lauenstein et al. applied this system for the detection of bone metastases in 26 patients, using coronal scans at five different levels with T1-weighted-gradient echo- , HASTE- and STIR sequences which were acquired within a total scan time of 40 min [8]. A high correlation was found between whole-body MRI and bone scintigraphy: 53 of 60 regions affected in bone scintigraphy were also identified in MRI. Although scintigraphy was again superior in detecting lesions in the rib cage and skull, MRI revealed additional lesions in the spine, femur and pelvis. In a most recently published study comparing whole-body MRI (3D-gradient echo data sets of five body levels) with CT, dedicated MRI and bone scintigraphy for the evaluation of distant metastases, whole-body MRI was most sensitive in the detection of hepatic and osseous metastases [32]. Whole-body MRI revealed osseous metastases in a total of 24 patients, scintigraphy revealed metastases in only 21 patients. 3.3. Parallel imaging and total imaging matrix (Tim) With the introduction of parallel imaging techniques (PAT), image acquisition time could be reduced [9–11]. Using this technique the image data is acquired simultane-
Fig. 1. Whole-body MRI with the use of a rolling platform. The rolling platform is mounted on top of the examination table, the body array fixed to the table is used for imaging of the whole body. Then the platform is moved step by step along the z-axis while the patient glides in between the body array and spine array-“sandwich”.
36
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
ously by two or more receiver coils with different spatial sensitivities and then reconstructed to an image. Image acquisition can be accelerated by reducing the number of acquired lines in the k-space at the costs of a lower signal-to-noise ratio. During the image reconstruction process, the missing data is generated by the spatial information of the different coil sensitivity profiles. This acquisition acceleration shortens the individual scan times, resulting in reduced overall examination times without compromises in spatial and temporal resolution and at a given examination time spatial resolution can even be further increased. A recently introduced 1,5 Tesla scanner (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany) combines 76 coil elements (“matrix coils”) and 32 receiver channels. It allows PAT in all three-dimensions with free table movement at a total field of view of 205 cm (Fig. 2a). The time reduction and faciliated setup makes it possible to make use of even potentially time consuming, but indispensible sequence types (e.g. STIR-sequences), for whole-body bone marrow imaging within an acceptable examination time. The protocol used for bone marrow screening (Fig. 2b) on this scanner incorporates T1-weighted-SE(TR 540, TE 13) and STIR-sequences (TR 2680, TE 101, TI 150). After one single positioning the patient is examined
at five body levels (head/neck, thorax/abdomen/arms, pelvis, proximal and distal lower extremities) with coronal STIRsequences (Fig. 3a). This is followed by coronal whole-body T1-weighted-SE sequences (Fig. 3b) and then sagittal STIRand T1-weighted imaging of the spine (Fig. 3c). The image data are automatically fused to one image with a special software without need of manual realignment, cropping or pasting. The total scanning time with this protocol is 43 min. First experiences at our institute with this new scanner show promising results in the detection of metastatic disease, especially to the bone: 20 patients were examined with both whole-body MRI and FDG-PET-CT for tumor staging [13]. FDG is a tracer that reflects an increased glucose metabolism in tumor tissue and is mainly used for tumor staging in patients with suspected visceral metastases and has proven to be very efficient in patients with lung tumors, malignant melanoma and tumors of the GI-tract [16,17]. Of 34 bone lesions which were depicted in PET-CT with a pathological FDG-uptake indicating malignancy, 29 were correctly identified in MRI, resulting in a sensitivity of 85% (Fig. 4). A reason for the lower performance could be the fact, that the lesions undected by MRI were mainly located in the ribcage. As in other comparative studies described before, this anatomical
Fig. 2. (a) Coil setup of a 32-channel whole-body MRI scanner (Magnetom Avanto) using 76 “matrix coils”: two head arrays, one neck array, two body arrays, one peripheral array and one spine array with combinable coil elements cover the whole bone allowing for parallel imaging in three-dimensions (Siemens Medical Solutions, Erlangen, Germany). (b) Examination protocol for whole-body bone marrow MRI screening using T1-weighted spin echo- and short tau inversion time inversion recovery (STIR) sequences.
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
37
Fig. 3. Fused T1w-spin echo- and STIR-images of the whole body in the coronal plane (a and b). T1w-spin echo- and STIR-images of the whole spine in the sagittal plane (Somatom Avanto, Siemens Medical Solutions, Erlangen) (c).
site may cause diagnostic difficulties, especially when wholebody sections in the coronal plane are acquired [6,8]. Due to the larger anatomic coverage of MRI (skull base to the proximal femoral bones in PET-CT), 10 additional malignant lesions were depicted in the appendicular skeleton (Fig. 5). The overall TNM stage in both PET-CT and MRI showed 90% concordance [13]. Antoch et al. analyzed the accuracy of both modalities in 98 patients with different oncological diseases in terms of TNM-based tumor staging. Both imaging procedures revealed a similar overall sensitvity in detecting distant metastases (PET/CT = 94%, MRI = 93%). Regarding skeletal metastases, the sensitivity was significantly higher when using whole-body MRI (85%) than PET-CT (62%) [18]. Recent study results indicate that MRI is superior to Fluoro-deoxyglucose (FDG)-PET in the assessment of skeletal metastases [15,25]. Thus, still only few studies exist which compare the performance of the dual modality approach with whole-body MRI for the detection of distant metastases [13,18].
4. Further indications for whole-body MR bone marrow imaging Considering the excellent performance of whole-body MRI in bone marrow examinations, this technique seems suitable for screening hemato-oncologic disorders.
Radiological diagnosis of multiple myeloma is based on plain radiography with X-ray images of the skull, spine, pelvis, humeral and femoral bones, exposing the patients to significant radiation. Sensitivity of radiography in the detection of myeloma manifestations, however, is rather low [22]. Still, radiographic skeletal surveys are accepted as part of the Salmon and Durie-classification, which grades the disease into three stages. Selection of therapy is based upon this classification [21]. MRI enables direct visualization of focal or diffuse myeloma infiltration of the marrow with high sensitivity and specificity [26,27]. However, MR imaging protocols often do not include the skull, sternum and ribs which are areas with high amounts of red marrow and represent frequent sites of infiltration. Whole-body MR imaging may be able to increase sensitivity of MRI in multiple myeloma by detecting infiltration at these sites. Ghanem et al. examined 32 patients with multiple myeloma using whole-body turboSTIR-imaging. In 25/26 patients, MRI readily detected histologically proven bone marrow infiltration, compared to 17/26 patients with conventional imaging [23]. Even when a diffuse infiltration pattern is present, which can be difficult to interpret at an early stage, MRI still shows a good sensitivity and high specificity, especially when contrast media application with calculation of percentage increase is included [27,28]. Baur et al. showed, that an extended staging system, including the MRI-status, has a significant influence on patient prog-
38
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
Fig. 4. Coronal whole-body STIR images of a 29-year-old woman with breast cancer. Solitary liver metastasis (arrow) in (a). Sagittal T1-weisghted-SE (b) and STIR-images (c) show bone metastases in various lumbar vertebral bodies and within the sacrum (arrows). (d) Metastasis within the left iliac bone (arrow).
nosis and survival rates: using the “classic” staging system of Durie and Salmon without MRI, 25 of 77 patients would have been understaged, showing all the more the importance to incorporate the MRI bone marrow status into the therapy planning of this disease [28]. The diagnosis of malignant lymphoma of the bone in many cases is delayed due to unspecific clinical signs and inapparent findings on radiographs and osteolytic lesions on plain film can be found in only half of the cases [33]. Extraskeletal involvement significantly decreases patient survival. Therefore whole-body MRI is an ideal candidate for skeletal manifestations of lymphoma. Whole-body MRI has been used in assessing bone marrow infiltration and extramedullary involvement by lymphoma. Although the modality of choice used in primary staging of this malignancy is 18 F-FDGPET, whole-body MRI may represent an alternative, especially to bone- or 67 Ga-scinitigraphy: Iizuka-Mikami et al. examined 34 patients with Non-Hodgkin’s lymphoma using
whole-body MRI and detected significantly more malignant lesions with MRI (n = 89) than with bone- (n = 14) or 67 Gascintigraphy (n = 5) [29]. Whole-body MR imaging has also been employed for the detection of bone tumors in children and young adults. Daldrup-Link et al. scanned 39 patients from 2 to 19 years with bone malignancies such as Ewing’s-sarcoma or lymphoma, and reported higher sensitivities for MRI (82%) than for bone scintigraphy (71%), whereas sensitivity of FDGPET was highest (90%) [30]. Especially in younger patients the high cellularity of hematopoietic bone marrow, which exhibits a low signal in T1-weighted scans, may limit the detection of bone metastases. In the rare event of a malignancy occurring while pregnancy, whole-body MRI should be used in order to avoid the risk of exposing the fetus to ionizing radiation. In 15% of patients with skeletal metastases the primary tumor is not known. In these cases an extensive and costly
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
39
Fig. 5. Coronal STIR-image of a 61 year-old woman with a neuroendocrine tumor. Large lymph node metastasis in the mediastinum and left iliacal region (a). Metastases in both femoral bones (b). Sagittal T1-weighted SE- and STIR-sequences of the whole spine show a multifocal bone marrow infiltration (c and d).
stepwise multimodality tumor search is frequently employed in order to identify the primary. It has been reported that whole-body MR imaging may allow the detection of a primary tumor, e.g. in the thyroid, prostate or lung, with the same accuracy as with the multimodality approach [31].
[3] [4] [5]
5. Conclusion The skeletal system is a frequent site of metastatic disease and early detection is important therapeutical decisions and patients outcome. Usually, different imaging modalities are used for bone marrow screening. Bone scintigraphy is still commonly used for the detection of skeletal metastases. MRI has proven to be the imaging modality with the highest sensitivity to detect malignant bone marrow infiltration without exposing the patient to ionizing radiation. With the development of better software and hardware such as new whole-body scanners, restriction, especially long examination times, could be overcome. Further evaluation, including larger patient collectives, is mandatory for assessing the potential of whole-body MRI. First results indicate, that it is a very promising method.
[6]
[7]
[8]
[9]
[10]
[11]
[12]
References [13] [1] Ghanem N, Altehoefer C, H¨ogerle S, et al. Comparative diagnostic value and therapeutic relevance of magnetic resonance imaging and bone marrow scintigraphy in patients with metastatic solid tumors of the axial skeleton. Eur J Radiol 2002;43:256–61. [2] Dalrup-Link HE, Franzius C, Link TM, et al. Whole-body MR imaging for detection of bone metastases in children and young
[14]
adults: comparison with skeletal scintigraphy and FDG PET. AJR 2001;177:229–36. Rubens RD. Bone metastases: the clinical problem. Eur J Cancer 1998;34:210–3. Krishnamurthy GT, Tubis M, Hiss J, Blahd WH. Distibution pattern of metastatic bone disease. JAMA 1977;237:837–42. Coleman RE. Monitoring of bone metastases. Eur J Cancer 1998;34:252–9. Steinborn M, Heuck AF, Tiling R, Bruegel M, Gauger L, Reiser MF. Whole body bone marrow MRI in patients with metastatic disease to the skeletal system. J Comput Assist Tomogr 1999;23: 123–9. Barkhausen J, Quick HH, Lauenstein T, et al. Whole-body MR imaging in 30 s with real-time true FISP and a continously rolling table platform: feasability study. Radiology 2001;220:252–6. Lauenstein T, Freudenberg L, Goehde S, et al. Whole body MRI using a rolling table platform for the detection of bone metastases. Eur Radiol 2002;12:2091–9. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI. Magn Reson Med 1999;42:952–62. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997;38:591–603. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002;47:1202–10. Eustace S, Tello R, DeCarvalho V, et al. A comparison of wholebody turbo STIR MR imaging and planar 99m TC-methylene diphosphonate scintigraphy in the examination of patients with suspected skeletal metastases. AJR 1997;169:1655–61. Schmidt GP, Baur A, Schoenberg SO, Reiser MF. High resolution whole body MRI tumor staging using parallel imaging compared to PET-CT: initial experiences on a 32-channel-MRI-system. Radiology 2004;44:889–98. Edelstyn GA, Gillespie PJ, Grebbel FS. The radiological demonstration of osseous metastases: experimental observations. Clin Radiol 1967;18:158–62.
40
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
[15] Ghanem N, Altehoefer C, Winterer J, H¨ogerle S, Ghanem-Wiesel C, Langer M. Whole body MRI in staging cancer patients: A comparison with whole-body FDG-PET. Cancer Imaging 2002;3:18–9. [16] Reske SN, Kotzerke J. FDG-PET for clinical use. Results of the third German interdisciplinary consensus conference, “Onko-PET III”, 21 July and 19 September 2000. Eur J Nucl Med 2001;28:1707– 23. [17] Pelosi E, Messa C, Sironi S, et al. Value of integrated PET/CT for lesion localisation in cancer patients: a comparative study. Eur J Nucl Med Mol Imaging 2004;31:932–9. [18] Antoch G, Vogt FM, Freudenberg LS, et al. Whole-body dualmodality PET/CT and whole-body MRI for tumor staging in oncology. JAMA 2003;290:3199–206. [19] Walker R, Kessar P, Blanchard R, et al. Turbo STIR magnetic resonance imaging as a whole-body screening tool for metastases in patients with breast carcinoma: preliminary clinical experience. J Magn Reson Imaging 2000;11:343–50. [20] Mehta RC, Marks MP, Hinks RS, Glover GH, Enzmann DR. MR evaluation of vertebral metatases: T1-weighted short inversion time inversion recovery, fast spin echo, and inversion-recovery fast spinecho sequences. Am J Neuroradiol 1995;16:281–8. [21] Durie BGM, Salmon SE. A clinical staging system for multiple myeloma. (Correlation of measured meyloma cell mass with presenting clinical features, response to treatment and survival.). Cancer 1975;36:842–54. [22] Lecouvet FE, Malghem J, Michaux L, et al. Skeletal survey in advanced multiple myeloma: radiographic versus MR imaging survey. Br J Haematol 1999;106:35–9. [23] Ghanem NA, Bley T, Springer O, Sch¨afer O, Th¨url C, Langer M. Whole body MRI of bone marrow in patients with multiple myeloma and monoclonal gammopathy in comparison to plain films.. In: Proceedings of the 11th International Society Magnetron Resonance Med Toronto. 2003. p. 115. [24] Kantarci M, Polat P, Alper F, et al. Comparison of CT and MRI for the diagnosis recurrent esophageal carcinoma after operation. Dis Esophagus 2004;17(1):32–7. [25] Gallowitsch HJ, Kresnik E, Gasser J, et al. F-18 fluorodeoxyglucose positron-emission tomography in the diagnosis of tumor recurrence and metastases in the follow-up of patients with breast carcinoma: a comparison to conventional imaging. Invest Radiol 2003;38: 250–6. [26] Staebler A, Baur A, Bartl R, Munker R, Lamerz R, Reiser M. Contrast enhancement and quantitative signal analysis in MRI of multiple myeloma: assessment of focal and diffuse growth patterns in
[27]
[28]
[29]
[30]
[31]
[32]
[33] [34]
[35] [36]
[37]
[38]
[39] [40]
marrow correlated with biopsies and survival rates. Am J Radiol 1996;167:1029–36. Baur A, St¨abler A, Bartl R, Lamerz R, Scheidler J, Reiser M. MRI gadolinium enhancement of bone marrow: age related changes in normals and in diffuse neoplastic infiltration. Skeletal Radiol 1997;26:414–8. Baur A, St¨abler A, Nagel D, et al. Magnetic resonance imaging as a supplement for the clinical staging system of Durie and Salmon? Cancer 2002;95:1334–45. Iizuka-Mikami M, Nagai K, Yoshida K, et al. Detection of bone marrow and extramedullary involvement in patients with non-Hodgkin’s lymphoma by whole-body MRI: comparison with bone and 67 Ga scintigraphies. Eur Radiol 2004;14:74–81. Daldrup-Link HE, Franzius C, Link TM, et al. Whole-body MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. AJR 2001;177:229–36. Eustace S, Tello R, DeCarvalho V, Carey J, Melhem E, Yucel EK. Whole body turbo STIR MRI in unknown primary tumor detection. J Magn Reson Imaging 1998;8:751–3. Lauenstein TC, Goehde SC, Herborn CU, et al. Whole-body MR imaging: evaluation of patients for metastases. Radiology 2004;233:139–48. Duerr HR, Mueller PE, Hiller E, et al. Malignant lymphoma of bone. Arch Orthop Trauma Surg 2002;122:10–6. Ghanem N, Kelly T, Altehoefer C, et al. Whole-body MRI in comparison to skeletal scintigraphy for detection of skeletal metastases in patients with solid tumors. Radiology 2004;44:864–73. Vanel D, Dromain C, Tardivon A. MRI of bone marrow disorders. Eur Radiol 2000;10:224–9. Poitout D, Gaujoux G, Lempidakis M, et al. X-ray computed tomography or MRI in the assessment of bone tumor extension. Chirurgie 1991;117(5-6):488–90. Baur-Melnyk A, Reiser M. Staging of multiple myeloma with MRI: comparison to MSCT and conventional radiography. Radiologe 2004;44:874–81. Fochem K, Ogris E. Early recognition of bone metastases (comparative study between bone scintigraphy and X-ray examination). Acta Med Austriatica 1976;3:170–6. Rieden K, Lellig U. Computed tomography and conventional X-ray diagnosis of tumor bone lesions. Rontgenblatter 1989;42:13–8. Krahe T, Nicolas V, Ring S, Warmuth-Metz M, Koster O. Diagnostic evaluation of full X-ray pictures and computed tomography of bone tumors of the spine. R¨oFo 1989;150:13–9.
Whole-body MR imaging of bone marrow G.P. Schmidt ∗ , S.O. Schoenberg, M.F. Reiser, A. Baur-Melnyk Institute of Clinical Radiology, University Hospitals Munich/Grosshadern, LMU, Marchioninistr. 15, M¨unchen 81377, Germany Received 17 January 2005; received in revised form 20 January 2005; accepted 26 January 2005
Abstract In clinical routine, multimodality algorithms, including X-ray, computed tomography, scintigraphy and MRI, are used in case of suspected bone marrow malignancy. Skeletal scintigraphy is widely used to asses metastatic disease to the bone, CT is the technique of choice to assess criteria of osseous destruction and bone stability. MRI is the only imaging technique that allows direct visualization of bone marrow and its components with high spatial resolution. The combination of unenhanced T1-weighted-spin echo- and turbo-STIR-sequences have shown to be most useful for the detection of bone marrow abnormalities and are able to discriminate benign from malignant bone marrow changes. Originally, whole-body MRI bone marrow screening was performed in sequential scanning techniques of five body levels with time consuming coil rearrangement and repositioning of the patient. The introduction of a rolling platform mounted on top of a conventional MRI examination table faciliated whole-body MR imaging and, with the use of fast gradient echo, T1-weighted and STIR-imaging techniques, for the first time allowed whole-body imaging within less than one hour. With the development of parallel imaging techniques (PAT) in combination with global matrix coil concepts, acquisition time could be reduced substantially without compromises in spatial resolution, enabling the implementation of more complex and flexible examination protocols. Whole-body MRI represents a new alternative to the stepwise multimodality concept for the detection of metastatic disease, multiple myeloma and lymphoma of the bone with high diagnostic accuracy. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: MRI; Whole body; Parallel imaging; Bone marrow; Metastases
1. Introduction The skeletal system is a frequent target of metastatic disease and constitutes the third most common location of metastatic manifestations after the liver and the lung. The primary tumors with the highest incidence of skeletal metastases are prostate cancer in men and breast cancer in women, followed by thyroid, renal and lung carcinomas. These tumors account for approximately 80% of metastatic bone disease [3]. Early detection of bone metastases has an important impact on patient management and may contribute to prevent complications like pathological fractures. Magnetic resonance imaging is the only imaging technique that allows direct visualization of the bone marrow and its components. It is the most sensitive technique for the detection of bone marrow pathologies [1,2]. Due to its high ∗
Corresponding author. Tel.: +49 89 7095 0; fax: +49 89 7095 8832. E-mail address: [email protected] (G.P. Schmidt). 0720-048X/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2005.01.019
soft tissue contrast and spatial resolution, MRI has become the method of choice for bone marrow imaging [35]. The predominantly hematogenous spread of bone metastases explains its frequent distribution in active hematopoetic bone marrow, especially in the axial skeleton. However, it has been reported that up to 40% of bone metastases occur in the appendicular skeleton stressing the need for whole-body anatomic coverage [4]. 2. Diagnostic imaging modalities for whole-body bone marrow screening In clinical practice most commonly multimodality algorithms are used in case of suspected metastatic bone disease, including conventional X-ray, bone marrow- or skeletal scintigraphy, PET, CT and MRI. These examinations are performed in order to evaluate the presence and type of bone lesions, to assess their extent and localization and eventually guide potential biopsy. All these modalities
34
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
have different performances in terms of sensitivity and specificity. 2.1. MRI versus radiography/CT In patients with a positive bone scan or with local symptoms and pain, radiographs and CT are usually performed for verification. Plain radiographs have a low sensitivity in the detection of metastases compared to MRI or other imaging modalities [22,37–39]. This is due to the fact that a change of 30–50% in mineral density is needed before a metastatic lesion becomes visible on plain film [14]. Compared to bone scintigraphy, it has been reported that radiography can be inapparent in nearly half of true-positive lesions revealed in scintigraphy [38]. Lecouvet et al. compared the performance of MRI and radiography to detect focal bone lesions in advanced Multiple myeloma and showed a clear superiority of MRI in given anatomic areas, like the spine (76% versus 42% detected lesions) and pelvis (75% versus 46% detected lesions) [22]. In the detection of bone destructions CT is far more sensitive than radiography: Krahe et al. compared conventional radiographs and CT examinations of 112 patients with primary and secondary bone tumors of the spine [40]. Of 268 involved vertebrae identified in CT, 88% were identified by radiographs when the vertebral body was infiltrated, but only 66% when other parts of the vertebrae were affected. Of the intraspinal and paravertebral tumor extensions only 23% and 33% were correctly diagnosed by plain radiography. Moreover, CT also allows to assess paraosseous tumor extension yields and it is the image modality of choice to evaluate the extent of osteolysis and to assess stability and fracture risk, respectively [24,36,39]. However, it is not yet clear if multislice-CT is equal to MRI for the assessment of bony metastases. 2.2. MRI versus scintigraphy In asymptomatic patients with suspected skeletal metastases bone scintigraphy is well established. This widely available technique is based on “indirectly” detecting bone metastases by registration of tracer uptake in areas of osteoblastic or osteoclastic activity and has long been regarded as a standard of reference [5]. In various studies, the efficiency of MRI has been compared to bone scintigraphy or bone marrow scintigraphy for the detection of bony metastases [1,2,6,12,34]. In a most recent study Ghanem et al. examined 129 patients with malignant tumors and suspected bone metastases using both whole-body MRI (coronary turbo-STIR-imaging) and bone scintigraphy [34]. Whole-body MRI and skeletal scintigraphy demonstrated skeletal metastases in 49/129 patients with a concordance of 81%. In this study, whole-body MRI revealed skeletal metastases in 15 cases which had not been found in skeletal scintigraphy. Moreover, MRI showed a larger extent of bone infiltration. In 60% of the cases, whole-body MRI evidenced tumor-associated findings, such as lymph node metastases in 42% of cases or metastases in
parenchymal organs in 23%. The same study group compared the performance of bone marrow scintigraphy with wholebody MRI in 20 patients with skeletal metastases: in 60% of patients MRI revealed more metastases and local therapy (e.g. radiotherapy or surgery) was applied on 78% of these patients [1]. The superiority of MRI compared to skeletal and bone marrow scintigraphy in depicting bone metastases has also been reported by other authors [2,6,12]. The high performance of MRI is mainly due to the fact that MRI can visually depict metastatic bone marrow infiltration at an early stage, before osseous changes occur due to osteoblastic or osteoclastic activity. For an MRI bone screening, the combination of non-contrasted T1-weighted-Spinechoand Turbo-STIR-sequences proved to be most sensitive and it also allows to reliably discriminate benign from malignant marrow disorders [6,19,20]. On T1-weighted sequences tumor spread is identified by replacement of normal marrow, resulting in an isointense or hypointense signal compared to muscle tissue. On fat suppressed sequences, like STIR (short tau inversion time inversion recovery), neoplastic lesions are readily detected by virtue of the hyperintense signal due to an increased content of water within tumor cells, in contrast to the surrounding normal marrow. In osteoblastic metastases areas of low signal intensity on T1-weighted turbo-SE images correspond to areas of low signal intensity on T2weighted turbo-SE images. On STIR-sequences osteoblastic metastases show a spectrum from no signal change in very dense sclerotic metastases to hyperintense signal in cases where more cellular components do exist [35]. Especially bone marrow infiltration by tumor cells which is not associated with bone destruction or formation of new bone, can be detected with MRI. Moreover, MRI also enables to differentiate reactive sclerosis, e.g. following fractures, from neoplastic osteosclerosis. The unique soft-tissue contrast of MRI enables for precise assessment of tumor infiltration within the bone marrow and adjacent paraosseous structures, such as the spinal canal. Finally, MRI can help to define the localization of biopsy, if needed. MRI also is particularly helpful to monitor therapy response, e.g. after radiation therapy [28].
3. Methods for whole-body MR Imaging The main challenge for whole-body MRI has been to cope with substantial differences in coil requirements, slice positioning and orientation into one comprehensive scan. Repositioning of the patient as well as the coils has been time consuming. Therefore, attempts for whole-body MRI were always associated with compromises in spatial resolution and image quality. 3.1. Sequential scanning In order to cover the whole body with a scanner of conventional design (Magnetom Symphony, Siemens Medical Solutions) five levels had to be scanned separately. First, the
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
patient was examined head first with coronal scans of the skull/neck and the thorax using a combination of head coil, neck array coil and 1–2 body array coils. Sagittal images of the spine were acquired using the spine array coil. T1weighted-spinecho- (TR 92,2, TE 4,1) and STIR-sequences (TR 3912, TE 60, TI 150) with a maximum field of view of 450 mm were obtained. After repositioning the patient feet first, the pelvis, femoral bones and lower legs were imaged in coronal planes, using body array and spine array coils and the same sequence types. Due to the need to reposition the patient, the total room time for this protocol was approximately 60 min. Steinborn et al. examined 18 patients with suspected skeletal metastases and correlated results with bone scintigraphy [6]. Out of 105 proven metastatic lesions, 96 (91.4%) were detected with whole-body MRI, whereas 89 lesions (84.8%) were detected with bone scintigraphy. In addition, MRI allowed to detect liver and lung metastases which of course, were not depicted in the bone scans. Approximately 1/4 of confirmed malignant lesions (23.8%) would have been missed in a “classic” MRI screening protocol covering only the spine, pelvis and proximal femurs. Nine skeletal metastases detected in the bone scan but missed in MRI were located either in the ribcage or in the skull, so that it can be conchided that these anatomical areas are not well represented with the described whole MRI technique. 3.2. Rolling table platform The introduction of dedicated rolling platforms with a fixed surface coil was a first step towards performing wholebody MRI within a single examination and towards a significant reduction of examination times and an increase in patient comfort [7]. The setup consists of a rolling table (length 270 cm, width 33–50 cm) mounted on top of the scanner table allowing free movement in the z-direction. For signal acquisition, the spine array and body phased array, which remains fixed to the scanner in the centre of the gantry, are employed. With the use of the rolling platform, the patient can be ex-
35
amined from head to toe sliding in between this so called “coil sandwich” (Fig. 1). Barkhausen et al. presented this setup for the first time as a quick whole-body MRI screening method for metastatic disease. The protocol was based on the acquisition of axial real-time trueFISP-gradient echo sequences (TR 2,2, TE 1,1; flip angle 60%). Using echo sharing to improve temporal resolution, the total scanning time was brought down to 30 s [7]. All 11 bone metastases detected by CT as a reference method were identified with MRI. Thus, four lesions depicted in bone scintigraphy, mainly located in the ribcage, were missed with MRI. A disadvantage of rapid acquisition data sets, which are often derived from gradient echo- or echoplanar imaging techniques, is a high sensitivity to image distortion caused by magnetic susceptibility effects, especially at air/soft-tissue interfaces as well as poor soft tissue contrast. Lauenstein et al. applied this system for the detection of bone metastases in 26 patients, using coronal scans at five different levels with T1-weighted-gradient echo- , HASTE- and STIR sequences which were acquired within a total scan time of 40 min [8]. A high correlation was found between whole-body MRI and bone scintigraphy: 53 of 60 regions affected in bone scintigraphy were also identified in MRI. Although scintigraphy was again superior in detecting lesions in the rib cage and skull, MRI revealed additional lesions in the spine, femur and pelvis. In a most recently published study comparing whole-body MRI (3D-gradient echo data sets of five body levels) with CT, dedicated MRI and bone scintigraphy for the evaluation of distant metastases, whole-body MRI was most sensitive in the detection of hepatic and osseous metastases [32]. Whole-body MRI revealed osseous metastases in a total of 24 patients, scintigraphy revealed metastases in only 21 patients. 3.3. Parallel imaging and total imaging matrix (Tim) With the introduction of parallel imaging techniques (PAT), image acquisition time could be reduced [9–11]. Using this technique the image data is acquired simultane-
Fig. 1. Whole-body MRI with the use of a rolling platform. The rolling platform is mounted on top of the examination table, the body array fixed to the table is used for imaging of the whole body. Then the platform is moved step by step along the z-axis while the patient glides in between the body array and spine array-“sandwich”.
36
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
ously by two or more receiver coils with different spatial sensitivities and then reconstructed to an image. Image acquisition can be accelerated by reducing the number of acquired lines in the k-space at the costs of a lower signal-to-noise ratio. During the image reconstruction process, the missing data is generated by the spatial information of the different coil sensitivity profiles. This acquisition acceleration shortens the individual scan times, resulting in reduced overall examination times without compromises in spatial and temporal resolution and at a given examination time spatial resolution can even be further increased. A recently introduced 1,5 Tesla scanner (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany) combines 76 coil elements (“matrix coils”) and 32 receiver channels. It allows PAT in all three-dimensions with free table movement at a total field of view of 205 cm (Fig. 2a). The time reduction and faciliated setup makes it possible to make use of even potentially time consuming, but indispensible sequence types (e.g. STIR-sequences), for whole-body bone marrow imaging within an acceptable examination time. The protocol used for bone marrow screening (Fig. 2b) on this scanner incorporates T1-weighted-SE(TR 540, TE 13) and STIR-sequences (TR 2680, TE 101, TI 150). After one single positioning the patient is examined
at five body levels (head/neck, thorax/abdomen/arms, pelvis, proximal and distal lower extremities) with coronal STIRsequences (Fig. 3a). This is followed by coronal whole-body T1-weighted-SE sequences (Fig. 3b) and then sagittal STIRand T1-weighted imaging of the spine (Fig. 3c). The image data are automatically fused to one image with a special software without need of manual realignment, cropping or pasting. The total scanning time with this protocol is 43 min. First experiences at our institute with this new scanner show promising results in the detection of metastatic disease, especially to the bone: 20 patients were examined with both whole-body MRI and FDG-PET-CT for tumor staging [13]. FDG is a tracer that reflects an increased glucose metabolism in tumor tissue and is mainly used for tumor staging in patients with suspected visceral metastases and has proven to be very efficient in patients with lung tumors, malignant melanoma and tumors of the GI-tract [16,17]. Of 34 bone lesions which were depicted in PET-CT with a pathological FDG-uptake indicating malignancy, 29 were correctly identified in MRI, resulting in a sensitivity of 85% (Fig. 4). A reason for the lower performance could be the fact, that the lesions undected by MRI were mainly located in the ribcage. As in other comparative studies described before, this anatomical
Fig. 2. (a) Coil setup of a 32-channel whole-body MRI scanner (Magnetom Avanto) using 76 “matrix coils”: two head arrays, one neck array, two body arrays, one peripheral array and one spine array with combinable coil elements cover the whole bone allowing for parallel imaging in three-dimensions (Siemens Medical Solutions, Erlangen, Germany). (b) Examination protocol for whole-body bone marrow MRI screening using T1-weighted spin echo- and short tau inversion time inversion recovery (STIR) sequences.
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
37
Fig. 3. Fused T1w-spin echo- and STIR-images of the whole body in the coronal plane (a and b). T1w-spin echo- and STIR-images of the whole spine in the sagittal plane (Somatom Avanto, Siemens Medical Solutions, Erlangen) (c).
site may cause diagnostic difficulties, especially when wholebody sections in the coronal plane are acquired [6,8]. Due to the larger anatomic coverage of MRI (skull base to the proximal femoral bones in PET-CT), 10 additional malignant lesions were depicted in the appendicular skeleton (Fig. 5). The overall TNM stage in both PET-CT and MRI showed 90% concordance [13]. Antoch et al. analyzed the accuracy of both modalities in 98 patients with different oncological diseases in terms of TNM-based tumor staging. Both imaging procedures revealed a similar overall sensitvity in detecting distant metastases (PET/CT = 94%, MRI = 93%). Regarding skeletal metastases, the sensitivity was significantly higher when using whole-body MRI (85%) than PET-CT (62%) [18]. Recent study results indicate that MRI is superior to Fluoro-deoxyglucose (FDG)-PET in the assessment of skeletal metastases [15,25]. Thus, still only few studies exist which compare the performance of the dual modality approach with whole-body MRI for the detection of distant metastases [13,18].
4. Further indications for whole-body MR bone marrow imaging Considering the excellent performance of whole-body MRI in bone marrow examinations, this technique seems suitable for screening hemato-oncologic disorders.
Radiological diagnosis of multiple myeloma is based on plain radiography with X-ray images of the skull, spine, pelvis, humeral and femoral bones, exposing the patients to significant radiation. Sensitivity of radiography in the detection of myeloma manifestations, however, is rather low [22]. Still, radiographic skeletal surveys are accepted as part of the Salmon and Durie-classification, which grades the disease into three stages. Selection of therapy is based upon this classification [21]. MRI enables direct visualization of focal or diffuse myeloma infiltration of the marrow with high sensitivity and specificity [26,27]. However, MR imaging protocols often do not include the skull, sternum and ribs which are areas with high amounts of red marrow and represent frequent sites of infiltration. Whole-body MR imaging may be able to increase sensitivity of MRI in multiple myeloma by detecting infiltration at these sites. Ghanem et al. examined 32 patients with multiple myeloma using whole-body turboSTIR-imaging. In 25/26 patients, MRI readily detected histologically proven bone marrow infiltration, compared to 17/26 patients with conventional imaging [23]. Even when a diffuse infiltration pattern is present, which can be difficult to interpret at an early stage, MRI still shows a good sensitivity and high specificity, especially when contrast media application with calculation of percentage increase is included [27,28]. Baur et al. showed, that an extended staging system, including the MRI-status, has a significant influence on patient prog-
38
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
Fig. 4. Coronal whole-body STIR images of a 29-year-old woman with breast cancer. Solitary liver metastasis (arrow) in (a). Sagittal T1-weisghted-SE (b) and STIR-images (c) show bone metastases in various lumbar vertebral bodies and within the sacrum (arrows). (d) Metastasis within the left iliac bone (arrow).
nosis and survival rates: using the “classic” staging system of Durie and Salmon without MRI, 25 of 77 patients would have been understaged, showing all the more the importance to incorporate the MRI bone marrow status into the therapy planning of this disease [28]. The diagnosis of malignant lymphoma of the bone in many cases is delayed due to unspecific clinical signs and inapparent findings on radiographs and osteolytic lesions on plain film can be found in only half of the cases [33]. Extraskeletal involvement significantly decreases patient survival. Therefore whole-body MRI is an ideal candidate for skeletal manifestations of lymphoma. Whole-body MRI has been used in assessing bone marrow infiltration and extramedullary involvement by lymphoma. Although the modality of choice used in primary staging of this malignancy is 18 F-FDGPET, whole-body MRI may represent an alternative, especially to bone- or 67 Ga-scinitigraphy: Iizuka-Mikami et al. examined 34 patients with Non-Hodgkin’s lymphoma using
whole-body MRI and detected significantly more malignant lesions with MRI (n = 89) than with bone- (n = 14) or 67 Gascintigraphy (n = 5) [29]. Whole-body MR imaging has also been employed for the detection of bone tumors in children and young adults. Daldrup-Link et al. scanned 39 patients from 2 to 19 years with bone malignancies such as Ewing’s-sarcoma or lymphoma, and reported higher sensitivities for MRI (82%) than for bone scintigraphy (71%), whereas sensitivity of FDGPET was highest (90%) [30]. Especially in younger patients the high cellularity of hematopoietic bone marrow, which exhibits a low signal in T1-weighted scans, may limit the detection of bone metastases. In the rare event of a malignancy occurring while pregnancy, whole-body MRI should be used in order to avoid the risk of exposing the fetus to ionizing radiation. In 15% of patients with skeletal metastases the primary tumor is not known. In these cases an extensive and costly
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
39
Fig. 5. Coronal STIR-image of a 61 year-old woman with a neuroendocrine tumor. Large lymph node metastasis in the mediastinum and left iliacal region (a). Metastases in both femoral bones (b). Sagittal T1-weighted SE- and STIR-sequences of the whole spine show a multifocal bone marrow infiltration (c and d).
stepwise multimodality tumor search is frequently employed in order to identify the primary. It has been reported that whole-body MR imaging may allow the detection of a primary tumor, e.g. in the thyroid, prostate or lung, with the same accuracy as with the multimodality approach [31].
[3] [4] [5]
5. Conclusion The skeletal system is a frequent site of metastatic disease and early detection is important therapeutical decisions and patients outcome. Usually, different imaging modalities are used for bone marrow screening. Bone scintigraphy is still commonly used for the detection of skeletal metastases. MRI has proven to be the imaging modality with the highest sensitivity to detect malignant bone marrow infiltration without exposing the patient to ionizing radiation. With the development of better software and hardware such as new whole-body scanners, restriction, especially long examination times, could be overcome. Further evaluation, including larger patient collectives, is mandatory for assessing the potential of whole-body MRI. First results indicate, that it is a very promising method.
[6]
[7]
[8]
[9]
[10]
[11]
[12]
References [13] [1] Ghanem N, Altehoefer C, H¨ogerle S, et al. Comparative diagnostic value and therapeutic relevance of magnetic resonance imaging and bone marrow scintigraphy in patients with metastatic solid tumors of the axial skeleton. Eur J Radiol 2002;43:256–61. [2] Dalrup-Link HE, Franzius C, Link TM, et al. Whole-body MR imaging for detection of bone metastases in children and young
[14]
adults: comparison with skeletal scintigraphy and FDG PET. AJR 2001;177:229–36. Rubens RD. Bone metastases: the clinical problem. Eur J Cancer 1998;34:210–3. Krishnamurthy GT, Tubis M, Hiss J, Blahd WH. Distibution pattern of metastatic bone disease. JAMA 1977;237:837–42. Coleman RE. Monitoring of bone metastases. Eur J Cancer 1998;34:252–9. Steinborn M, Heuck AF, Tiling R, Bruegel M, Gauger L, Reiser MF. Whole body bone marrow MRI in patients with metastatic disease to the skeletal system. J Comput Assist Tomogr 1999;23: 123–9. Barkhausen J, Quick HH, Lauenstein T, et al. Whole-body MR imaging in 30 s with real-time true FISP and a continously rolling table platform: feasability study. Radiology 2001;220:252–6. Lauenstein T, Freudenberg L, Goehde S, et al. Whole body MRI using a rolling table platform for the detection of bone metastases. Eur Radiol 2002;12:2091–9. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI. Magn Reson Med 1999;42:952–62. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997;38:591–603. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002;47:1202–10. Eustace S, Tello R, DeCarvalho V, et al. A comparison of wholebody turbo STIR MR imaging and planar 99m TC-methylene diphosphonate scintigraphy in the examination of patients with suspected skeletal metastases. AJR 1997;169:1655–61. Schmidt GP, Baur A, Schoenberg SO, Reiser MF. High resolution whole body MRI tumor staging using parallel imaging compared to PET-CT: initial experiences on a 32-channel-MRI-system. Radiology 2004;44:889–98. Edelstyn GA, Gillespie PJ, Grebbel FS. The radiological demonstration of osseous metastases: experimental observations. Clin Radiol 1967;18:158–62.
40
G.P. Schmidt et al. / European Journal of Radiology 55 (2005) 33–40
[15] Ghanem N, Altehoefer C, Winterer J, H¨ogerle S, Ghanem-Wiesel C, Langer M. Whole body MRI in staging cancer patients: A comparison with whole-body FDG-PET. Cancer Imaging 2002;3:18–9. [16] Reske SN, Kotzerke J. FDG-PET for clinical use. Results of the third German interdisciplinary consensus conference, “Onko-PET III”, 21 July and 19 September 2000. Eur J Nucl Med 2001;28:1707– 23. [17] Pelosi E, Messa C, Sironi S, et al. Value of integrated PET/CT for lesion localisation in cancer patients: a comparative study. Eur J Nucl Med Mol Imaging 2004;31:932–9. [18] Antoch G, Vogt FM, Freudenberg LS, et al. Whole-body dualmodality PET/CT and whole-body MRI for tumor staging in oncology. JAMA 2003;290:3199–206. [19] Walker R, Kessar P, Blanchard R, et al. Turbo STIR magnetic resonance imaging as a whole-body screening tool for metastases in patients with breast carcinoma: preliminary clinical experience. J Magn Reson Imaging 2000;11:343–50. [20] Mehta RC, Marks MP, Hinks RS, Glover GH, Enzmann DR. MR evaluation of vertebral metatases: T1-weighted short inversion time inversion recovery, fast spin echo, and inversion-recovery fast spinecho sequences. Am J Neuroradiol 1995;16:281–8. [21] Durie BGM, Salmon SE. A clinical staging system for multiple myeloma. (Correlation of measured meyloma cell mass with presenting clinical features, response to treatment and survival.). Cancer 1975;36:842–54. [22] Lecouvet FE, Malghem J, Michaux L, et al. Skeletal survey in advanced multiple myeloma: radiographic versus MR imaging survey. Br J Haematol 1999;106:35–9. [23] Ghanem NA, Bley T, Springer O, Sch¨afer O, Th¨url C, Langer M. Whole body MRI of bone marrow in patients with multiple myeloma and monoclonal gammopathy in comparison to plain films.. In: Proceedings of the 11th International Society Magnetron Resonance Med Toronto. 2003. p. 115. [24] Kantarci M, Polat P, Alper F, et al. Comparison of CT and MRI for the diagnosis recurrent esophageal carcinoma after operation. Dis Esophagus 2004;17(1):32–7. [25] Gallowitsch HJ, Kresnik E, Gasser J, et al. F-18 fluorodeoxyglucose positron-emission tomography in the diagnosis of tumor recurrence and metastases in the follow-up of patients with breast carcinoma: a comparison to conventional imaging. Invest Radiol 2003;38: 250–6. [26] Staebler A, Baur A, Bartl R, Munker R, Lamerz R, Reiser M. Contrast enhancement and quantitative signal analysis in MRI of multiple myeloma: assessment of focal and diffuse growth patterns in
[27]
[28]
[29]
[30]
[31]
[32]
[33] [34]
[35] [36]
[37]
[38]
[39] [40]
marrow correlated with biopsies and survival rates. Am J Radiol 1996;167:1029–36. Baur A, St¨abler A, Bartl R, Lamerz R, Scheidler J, Reiser M. MRI gadolinium enhancement of bone marrow: age related changes in normals and in diffuse neoplastic infiltration. Skeletal Radiol 1997;26:414–8. Baur A, St¨abler A, Nagel D, et al. Magnetic resonance imaging as a supplement for the clinical staging system of Durie and Salmon? Cancer 2002;95:1334–45. Iizuka-Mikami M, Nagai K, Yoshida K, et al. Detection of bone marrow and extramedullary involvement in patients with non-Hodgkin’s lymphoma by whole-body MRI: comparison with bone and 67 Ga scintigraphies. Eur Radiol 2004;14:74–81. Daldrup-Link HE, Franzius C, Link TM, et al. Whole-body MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. AJR 2001;177:229–36. Eustace S, Tello R, DeCarvalho V, Carey J, Melhem E, Yucel EK. Whole body turbo STIR MRI in unknown primary tumor detection. J Magn Reson Imaging 1998;8:751–3. Lauenstein TC, Goehde SC, Herborn CU, et al. Whole-body MR imaging: evaluation of patients for metastases. Radiology 2004;233:139–48. Duerr HR, Mueller PE, Hiller E, et al. Malignant lymphoma of bone. Arch Orthop Trauma Surg 2002;122:10–6. Ghanem N, Kelly T, Altehoefer C, et al. Whole-body MRI in comparison to skeletal scintigraphy for detection of skeletal metastases in patients with solid tumors. Radiology 2004;44:864–73. Vanel D, Dromain C, Tardivon A. MRI of bone marrow disorders. Eur Radiol 2000;10:224–9. Poitout D, Gaujoux G, Lempidakis M, et al. X-ray computed tomography or MRI in the assessment of bone tumor extension. Chirurgie 1991;117(5-6):488–90. Baur-Melnyk A, Reiser M. Staging of multiple myeloma with MRI: comparison to MSCT and conventional radiography. Radiologe 2004;44:874–81. Fochem K, Ogris E. Early recognition of bone metastases (comparative study between bone scintigraphy and X-ray examination). Acta Med Austriatica 1976;3:170–6. Rieden K, Lellig U. Computed tomography and conventional X-ray diagnosis of tumor bone lesions. Rontgenblatter 1989;42:13–8. Krahe T, Nicolas V, Ring S, Warmuth-Metz M, Koster O. Diagnostic evaluation of full X-ray pictures and computed tomography of bone tumors of the spine. R¨oFo 1989;150:13–9.